MRI Fusion Research Articles

P025 Imaging and AI supported transformation to better practice - MRI fusion target boost LDR prostate brachytherapy

P025 Imaging and AI supported transformation to better practice - MRI fusion target boost LDR prostate brachytherapy

Pathologic prostate cancer grade concordance among high-resolution micro-ultrasound, systematic transrectal ultrasound and MRI fusion biopsy

Background and ObjectiveComparative studies among biopsy strategies have not been conducted evaluating pathologic concordance at radical prostatectomy(RP), especially with novel micro-ultrasound (micro-US) image-guided biopsy. MethodsA retrospective study among patients with PCa who underwent RP following TRUS, MRI-TRUS fusion, microUS, or MRI-microUS fusion biopsy in a multi-site single institution. We compared GG-upgrade from biopsy to RP based on highest GG in any biopsy core and examined clinical/pathologic factors associated with pathologic upgrading using descriptive statistics, and multivariable logistic-regression analysis. Results429 patients between 1/2021 and 6/2023 including 10 (25.6%) who underwent systematic TRUS, 237 (55.2%) MRI-TRUS, 67 (15.6%) MRI-microUS and 15 (3.5%) micoUS-alone biopsy prior to RP. 78 (18.2%) were upgraded on final pathology (TRUS 31 (28.2%), MRI-TRUS 31 (13.1%), MRI-microUS 10 (14.9%), microUS: 6 (40%)) and 99 downgraded. 14 (3.5%) experienced a major upgrade (≥2 GG increase). On multivariable-analysis both MRI-TRUS (odds ratio, OR: 0.31,95% CI:0.17–0.56, P < 0.001) and MRI-microUS (OR: 0.43,95%CI: 0.19–0.98, P = 0.044) were associated with lower odds pathological-upgrade compared with TRUS biopsy alone. No significant differences in the odds of upgrade between TRUS and microUS alone (P > 0.05), or between MRI-microUS and MRI-TRUS(P = 0.696) on pairwise comparisons. MRI-microUS was associated with lower upgrade compared with microUS (OR: 0.26,95% CI:0.08–0.90, P = 0.034). No difference among the biopsy strategies in pathologic downgrading or overall GG concordance. Limitations include retrospective analysis, inter-clinician experience and lesion selection in varying biopsy techniques. ConclusionBoth MRI-microUS and MRI-TRUS fusion were associated with similarly improved GG concordance compared with TRUS biopsy. No significant differences between microUS-alone and TRUS or between MRI-microUS and MRI-TRUS fusion approaches, may suggest similar accuracy performance for disease sampling. What does the study addTo our knowledge, this is the first study to investigate GG concordance based on type of biopsy, especially microUS related GG upgrading after RP. In a moderately sized cohort this is the first to investigate pathologic concordance in MRI-microUS fusion compared to MRI-TRUS fusion biopsy. Our study may help urologists in counseling patients after biopsy and choosing the ideal image guided biopsy technique, however randomized controlled trials are needed to validate our results. Patient SummaryWe performed a study to see if the type of prostate biopsy, including use of MRI assistance as well as a new image-guided biopsy using a more advanced ultrasound, was better able to identify the aggressiveness of prostate cancer patients had. We found that the new biopsy type when fused with MRI and the existing MRI-guided biopsy type were similar in predicting the type of prostate cancer found at prostate surgery. These were both more accurate than the conventional ultrasound only biopsy.

Study on the classification of benign and malignant breast lesions using a multi-sequence breast MRI fusion radiomics and deep learning model

PurposeTo develop a multi-modal model combining multi-sequence breast MRI fusion radiomics and deep learning for the classification of benign and malignant breast lesions, to assist clinicians in better selecting treatment plans. MethodsA total of 314 patients who underwent breast MRI examinations were included. They were randomly divided into training, validation, and test sets in a ratio of 7:1:2. Subsequently, features of T1-weighted images (T1WI), T2-weighted images (T2WI), and dynamic contrast-enhanced MRI (DCE-MRI) were extracted using the convolutional neural network ResNet50 for fusion, and then combined with radiomic features from the three sequences. The following models were established: T1 model, T2 model, DCE model, DCE_T1_T2 model, and DCE_T1_T2_rad model. The performance of the models was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, specificity, positive predictive value, and negative predictive value. The differences between the DCE_T1_T2_rad model and the other four models were compared using the Delong test, with a P-value < 0.05 considered statistically significant. ResultsThe five models established in this study performed well, with AUC values of 0.53 for the T1 model, 0.62 for the T2 model, 0.79 for the DCE model, 0.94 for the DCE_T1_T2 model, and 0.98 for the DCE_T1_T2_rad model. The DCE_T1_T2_rad model showed statistically significant differences (P < 0.05) compared to the other four models. ConclusionThe use of a multi-modal model combining multi-sequence breast MRI fusion radiomics and deep learning can effectively improve the diagnostic performance of breast lesion classification.

Inter-reader reliability and diagnostic accuracy of PI-RADS scoring between academic and community care networks: How wide is the gap?

ImportanceThe Prostate Imaging Reporting & Data System (PI-RADS) scoring guidelines were developed to address the substantial variation in interpretation and reporting of prostate cancer (PCa) multiparametric MRI (mpMRI) results, and subsequent updates have sought to further improve inter-reader reliability. Nonetheless, the variability of PI-RADS scoring in real-world settings may represent a continuing challenge to the widespread standardization of prostate mpMRI and limit its overall clinical benefit. ObjectiveTo assess variability in mpMRI interpretation and reporting of PCa, we evaluated the discrepancies in PI-RADS scoring between community practices and a tertiary academic care center. Design, setting, and participantsWe identified 262 mpMRI studies from nonacademic facilities, reinterpreted by radiologists at our institution between January 2016 and July 2022. Results of targeted MRI fusion biopsy were identified for 193 of these patients, totaling 302 lesions. PI-RADS scoring from both community and academic interpreters were recorded in addition to presence of clinically significant PCa (csPCa) on pathological analysis of targeted cores. Main Outcome and measuresThe primary outcome was inter-reader reliability via intraclass correlation (ICC) and the kappa statistic. We also assessed the diagnostic accuracy of PI-RADS scoring for detecting csPCa for both cohorts via receiver operator characteristics (ROC) analysis and compared these findings using paired-sample area difference under curve analysis. ResultsInter-reader agreement and reliability of PI-RADS scoring per lesion was generally poor (absolute agreement ICC = 0.393, 95% CI: 0.288–0.488; consistency ICC = 0.407, 95% CI: 0.308–0.497; kappa = 0.336, 95% CI: 0.267–0.406). Reliability results from studies obtained after the publication of PI-RADSv2.1 were similar to those of the overall analysis. No agreement was observed in the subgroup of lesions scored as PIRADS 3 by community interpreters. No statistically significant difference in diagnostic accuracy was observed between cohorts (ROC area under curve [AUC]: 0.759 vs. 0.785, respectively; P = 0.337). PI-RADS 3 was determined to be the optimal cutoff for detecting clinically significant disease in both cohorts. Conclusions and relevanceOur results suggest that mpMRI diagnostic accuracy for detecting csPCa is not significantly different between academic and community practices. However, significantly poor reliability of mpMRI was observed between cohorts, suggesting the risk of introducing practice variation for community prostate cancer management. Variability, particularly for PI-RADS 3 lesions, can lead to inconsistent biopsy recommendations, which may result in missed csPCa or unnecessary biopsies.

Is YouTube really a source of information on MRI guided prostate biopsy?

Background: The purpose of this review was to evaluate the overall quality of the information in YouTube™ videos about prostate MRI -guided biopsy and whether it describes the procedure. Materials and methods: The terms " MRI -guided biopsy of the prostate" and "MRI targeted prostate biopsy" were searched on YouTube. The first 118 videos of the results were selected, and the written information of the British Association of Urological Surgeons for transrectal ultrasound-guided MRI fusion biopsy was performed. For MRI-guided prostate biopsy, four more questions were added to the MRI-guided prostate biopsy criteria and reviewed by two authors. A qualitative rating was given based on how well they provided information about factors such as preparation for the procedure, the mechanism of the procedure, and possible side effects. In addition, who published the videos (doctor, patient, or YouTuber), target audience (doctors or patients), video duration, upload date, number of daily views, total number of views, number of likes, and comments were recorded. The reliability and quality of the content of these videos included in the study were evaluated using JAMA and GQS scores. Results: A total of 118 videos were reviewed; 3 videos were rated "excellent" and 56 videos were rated "very poor". 91 of the videos (77%) were prepared for physicians, and the remaining videos were prepared by non-physicians. The average duration of the videos was 9.8 minutes, and the average number of daily views was 14.3. The number of views varied between 19 and 710,399 (average: 21). While the average of likes was 62.5, the average of comments was 4.3. The average number of views was 4.3. The score was found to be 3.54 ± 0.5 out of 5. The GQS average was calculated as 4.42 ± 0.56 out of 5. When the videos uploaded by physicians and non-physicians were compared, no statistically significant difference was found in terms of the average duration of the videos, JAMA scores, or number of comments. The number of daily views, number of likes, total number of views, GQS, and scoring criteria were statistically higher in the videos prepared by non-physicians. Conclusion: Information about MRI-guided prostate biopsies on YouTube is not of a high enough standard to allow patients to make informed decisions. Therefore, healthcare professionals must produce high-quality, informative, patient-focused medical YouTube videos.

CT and MRI Image Fusion via Coupled Feature-Learning GAN

The fusion of multimodal medical images, particularly CT and MRI, is driven by the need to enhance the diagnostic process by providing clinicians with a single, comprehensive image that encapsulates all necessary details. Existing fusion methods often exhibit a bias towards features from one of the source images, making it challenging to simultaneously preserve both structural information and textural details. Designing an effective fusion method that can preserve more discriminative information is therefore crucial. In this work, we propose a Coupled Feature-Learning GAN (CFGAN) to fuse the multimodal medical images into a single informative image. The proposed method establishes an adversarial game between the discriminators and a couple of generators. First, the coupled generators are trained to generate two real-like fused images, which are then used to deceive the two coupled discriminators. Subsequently, the two discriminators are devised to minimize the structural distance to ensure the abundant information in the original source images is well-maintained in the fused image. We further empower the generators to be robust under various scales by constructing a discriminative feature extraction (DFE) block with different dilation rates. Moreover, we introduce a cross-dimension interaction attention (CIA) block to refine the feature representations. The qualitative and quantitative experiments on common benchmarks demonstrate the competitive performance of the CFGAN compared to other state-of-the-art methods.

Prior Negative Biopsy, PSA Density, and Anatomic Location Impact Cancer Detection Rate of MRI-Targeted PI-RADS Index Lesions.

MRI fusion prostate biopsy has improved the detection of clinically significant prostate cancer (CSC). Continued refinements in predicting the pre-biopsy probability of CSC are essential for optimal patient counseling. We investigated potential factors related to improved cancer detection rates (CDR) of CSC in patients with PI-RADS ≥ 3 lesions. The pathology of 980 index lesions in 980 patients sampled by transrectal mpMRI-targeted prostate biopsy across four medical centers between 2017-2020 was reviewed. PI-RADS lesion distribution included 291 PI-RADS-5, 374 PI-RADS-4, and 315 PI-RADS-3. We compared CDR of index PI-RADS ≥ 3 lesions based on location (TZ) vs. (PZ), PSA density (PSAD), and history of prior negative conventional transrectal ultrasound-guided biopsy (TRUS). Mean age, PSA, prostate volume, and level of prior negative TRUS biopsy were 66 years (43-90), 7.82 ng/dL (5.6-11.2), 54 cm3 (12-173), and 456/980 (46.5%), respectively. Higher PSAD, no prior history of negative TRUS biopsy, and PZ lesions were associated with higher CDR. Stratified CDR highlighted significant variance across subgroups. CDR for a PI-RADS-5 score, PZ lesion with PSAD ≥ 0.15, and prior negative biopsy was 77%. Conversely, the CDR rate for a PI-RADS-4 score, TZ lesion with PSAD < 0.15, and prior negative biopsy was significantly lower at 14%. For index PI-RADS ≥ 3 lesions, CDR varied significantly based on location, prior history of negative TRUS biopsy, and PSAD. Such considerations are critical when counseling on the merits and potential yield of prostate needle biopsy.

Effect of polygenic risk score for clinically significant prostate cancer in a screening program: The BARCODE 1 study results.

10500 Background: Incidence of prostate cancer (PCa) is increasing, but there is no internationally agreed population screening program. Studies using an age-based PSA approach show a high rate of false-positive results as well as over-diagnosis of indolent PCa. Genome wide association studies identify common germline variants to calculate a polygenic risk score (PRS) associated with PCa risk. The BARCODE1 study used PRS to target PCa screening to those at higher risk based on genotype. Methods: European men aged 55-69yrs were recruited via Primary Care in the UK. PRS was constructed by summing weighted risk alleles for 130 PCa risk variants using germline DNA from saliva samples via mailed kits. Men with a PRS &gt; 90th centile were invited for PCa screening using MRI and 12-core transperineal biopsy (including MRI fusion to target additional lesions where identified) irrespective of PSA result. Results: Invitation letters were sent to 40,292 men. 8,953 (22%) expressed an interest; 8,014 were eligible and sent a saliva kit. 6,644 consented; 6,393 were genotyped; 251 failed QC. A total of 6,142 participants had PRS calculated: 745 (12.1%) had a PRS &gt; 90th centile and were invited to screening. 558/745 participants attended screening (121 declined, 66 excluded on health grounds). 551 underwent MRI and 468 had prostate biopsy resulting in 187 (40.0%) diagnoses of PCa, overall PCa detection rate 2.8%. Mean age at diagnosis 64.1yrs (range 57-73; median 64). Using NCCN criteria (2023) 103/187 (55.1%) of cancers were Intermediate or High Risk; 40/187 (21.4%) were Intermediate Unfavourable/High/Very High Risk. 119/187 (63.6%) men had a PSA &lt;3.0ug/L; PPV of biopsy for PSA &gt; 3.0ug/L was 49.6%. PPV of MRI (presence of PI-RADS 3-5 lesion) 60.4%. PPV of PRS alone 40%. 103/187 (55.1%) had Gleason &gt;7; compared with 360/1014 (35.5%) p &lt; 0.001 in the PSA directed ERSPC study. Conclusions: A population PCa screening program using PRS risk-stratification enriches for clinically significant PCa requiring treatment. It detects a high proportion of clinically significant disease compared with PSA or MRI based screening programs and MRI missed a significant proportion (17-67%) of cancers found on biopsy. This is the first study to assess if this approach will be useful in population screening programs. Clinical trial information: NCT03857477 . [Table: see text]

A Hybrid Model for Fetal Growth Restriction Assessment by Automatic Placental Radiomics on T2-Weighted MRI and Multifeature Fusion.

MRI-based placental analyses have been used to improve fetal growth restriction (FGR) assessment by complementing ultrasound-based measurements. However, these are still limited by time-consuming manual annotation in MRI data and the lack of mother-based information. To develop and validate a hybrid model for accurate FGR assessment by automatic placental radiomics on T2-weighted imaging (T2WI) and multifeature fusion. Retrospective. 274 pregnant women (29.5 4.0 years) from two centers were included and randomly divided into training (N = 119), internal test (N = 40), time-independent validation (N = 43), and external validation (N = 72) sets. 1.5-T, T2WI half-Fourier acquisition single-shot turbo spin-echo pulse sequence. First, the placentas on T2WI were manually annotated, and a deep learning model was developed to automatically segment the placentas. Then, the radiomic features were extracted from the placentas and selected by three-step feature selection. In addition, fetus-based measurement features and mother-based clinical features were obtained from ultrasound examinations and medical records, respectively. Finally, a hybrid model based on random forest was constructed by fusing these features, and further compared with models based on other machine learning methods and different feature combinations. The performances of placenta segmentation and FGR assessment were evaluated by Dice similarity coefficient (DSC) and the area under the receiver operating characteristic curve (AUROC), respectively. A P-value <0.05 was considered statistically significant. The placentas were automatically segmented with an average DSC of 90.0%. The hybrid model achieved an AUROC of 0.923, 0.931, and 0.880 on the internal test, time-independent validation, and external validation sets, respectively. The mother-based clinical features resulted in significant performance improvements for FGR assessment. The proposed hybrid model may be able to assess FGR with high accuracy. Furthermore, information complementation based on placental, fetal, and maternal features could also lead to better FGR assessment performance. Stage 2.

The role of multiparametric magnetic resonance ımaging in the diagnosis of granulomatous prostatitis mimicking prostate cancer.

Aimed to investigate the role of multiparametric magnetic resonance imaging (mp-MRI) in the diagnosis of granulomatous prostatitis caused by intravesical Bacillus Calmette-Guérin (BCG). In this prospective, single-center study, 10 male patients who were given intravesical BCG due to intermediate- and high-risk bladder cancer were included. Before transurethral resection of bladder tumors (TURB), all patients were evaluated by mp-MRI, serum prostate-specific antigen (PSA), and digital rectal examination (DRE). Serum PSA levels and DRE findings were evaluated before and after intravesical BCG treatment. Prostate mp-MRI was performed for patients with elevated levels of serum PSA and/or with abnormal DRE findings. Then, MRI fusion + systematic prostate biopsy was performed. Demographic data of the patients before and after intravesical BCG were compared. The average age of the patients was 66.9years (55-87years). While PSA was 1.7ng/ml before intravesical BCG treatment, it was 4.3ng/ml after intravesical BCG treatment (p = 0.005). PSA density (PSAD) was 0.04 and 0.10 before and after the treatment, respectively (p = 0.012). DRE findings of all patients were normal before the treatment. However, abnormal findings were detected in 80% of them after the treatment (p = 0.008). PI-RADS ≥ 3 lesions were found to be significantly higher in all patients after intravesical BCG (p = 0.004). Granulomatous prostatitis is a rare complication of intravesical BCG. High PSA, abnormal DRE, and PI-RADS ≥ 3 lesions detected after intravesical BCG should suggest granulomatous prostatitis and unnecessary biopsies may be avoided.

Prostate cancer detection and complications of MRI-targeted prostate biopsy using cognitive registration, software-assisted image fusion or in-bore guidance: a systematic review and meta-analysis of comparative studies.

Three primary strategies for MRI-targeted biopsies (TB) are available: Cognitive TB (COG-TB), MRI-US Fusion TB (FUS-TB), and In Bore TB (IB-TB). Despite nearly a decade of practice, a consensus on the preferred approach is lacking, with previous studies showing comparable PCa detection rates among the three methods. We conducted a search of PubMed, EMBASE, PubMed, Web of Science, and Scopus databases from 2014 to 2023, to identify studies comparing at least two of the three methods and reporting clinically significant PCa (csPCa) detection rates. The primary and secondary outcomes were to compare the csPCa and insignificant prostate cancer (iPCa, ISUP GG 1) detection rates between TB techniques. The tertiary outcome was to compare the complication rate between TB techniques. Detection rates were pooled using random-effect models. Planned sensitivity analyses included subgroup analysis according to the definition of csPCa and positive MRI, previous biopsy status, biopsy route, prostate volume, and lesion characteristics. A total of twenty studies, involving 4928 patients, were included in the quantitative synthesis. The meta-analysis unveiled comparable csPCa detection rates among COG-TB (0.37), FUS-TB (0.39), and IB-TB (0.47). iPCa detection rate was also similar between TB techniques (COG-TB: 0.12, FUS-TB: 0.17, IB-TB: 0.18). All preplanned sensitivity analyses were conducted and did not show any statistically significant difference in the detection of csPCa between TB methods. Complication rates, however, were infrequently reported, and when available, no statistically significant differences were observed among the techniques. This unique study, exclusively focusing on comparative research, indicates no significant differences in csPCa and iPCa detection rates between COG-TB, FUS-TB, and IB-TB. Decisions between these techniques may extend beyond diagnostic accuracy, considering factors such as resource availability and operator preferences. Well-designed prospective studies are warranted to refine our understanding of the optimal approach for TB in diverse clinical scenarios.

Evaluation of 3D C-Arm Fluoroscopy versus Diagnostic CT for Deep Brain Stimulation Stereotactic Registration and Post-Operative Lead Localization

Introduction: DBS efficacy depends on accuracy. CT-MRI fusion is established for both stereotactic registration and electrode placement verification. The desire to streamline DBS workflows, reduce operative time, and minimize patient transfers has increased interest in portable imaging modalities such as the Medtronic O-arm® and mobile CT. However, these remain expensive and bulky. 3D C-arm fluoroscopy (3DXT) units are a smaller and less costly alternative, albeit incompatible with traditional frame-based localization and without useful soft tissue resolution. We aimed to compare fusion of 3DXT and CT with pre-operative MRI to evaluate if 3DXT-MRI fusion alone is sufficient for accurate registration and reliable targeting verification. We further assess DBS targeting accuracy using a 3DXT workflow and compare radiation dosimetry between modalities. Methods: Patients underwent robot-assisted DBS implantation using a workflow incorporating 3DXT which we describe. Two intra-operative 3DXT spins were performed for registration and accuracy verification followed by conventional CT post-operatively. Post-operative 3DXT and CT images were independently fused to the same pre-operative MRI sequence and co-ordinates generated for comparison. Registration accuracy was compared to 15 consecutive controls who underwent CT-based registration. Radial targeting accuracy was calculated and radiation dosimetry recorded. Results: Data were obtained from 29 leads in 15 consecutive patients. 3DXT registration accuracy was significantly superior to CT with mean error 0.22 ± 0.03 mm (p < 0.0001). Mean Euclidean electrode tip position variation for CT to MRI versus 3DXT to MRI fusion was 0.62 ± 0.40 mm (range 0.0 mm–1.7 mm). In comparison, direct CT to 3DXT fusion showed electrode tip Euclidean variance of 0.23 ± 0.09 mm. Mean radial targeting accuracy assessed on 3DXT was 0.97 ± 0.54 mm versus 1.15 ± 0.55 mm on CT with differences insignificant (p = 0.30). Mean patient radiation doses were around 80% lower with 3DXT versus CT (p < 0.0001). Discussion: Mobile 3D C-arm fluoroscopy can be safely incorporated into DBS workflows for both registration and lead verification. For registration, the limited field of view requires the use of frameless transient fiducials and is highly accurate. For lead position verification based on MRI co-registration, we estimate there is around a 0.4 mm discrepancy between lead position seen on 3DXT versus CT when corrected for brain shift. This is similar to that described in O-arm® or mobile CT series. For units where logistical or financial considerations preclude the acquisition of a cone beam CT or mobile CT scanner, our data support portable 3D C-arm fluoroscopy as an acceptable alternative with significantly lower radiation exposure.

A New Automated Prognostic Prediction Method Based on Multi-Sequence Magnetic Resonance Imaging for Hepatic Resection of Colorectal Cancer Liver Metastases.

Colorectal cancer is a prevalent and life-threatening disease, where colorectal cancer liver metastasis (CRLM) exhibits the highest mortality rate. Currently, surgery stands as the most effective curative option for eligible patients. However, due to the insufficient performance of traditional methods and the lack of multi-modality MRI feature complementarity in existing deep learning methods, the prognosis of CRLM surgical resection has not been fully explored. This paper proposes a new method, multi-modal guided complementary network (MGCNet), which employs multi-sequence MRI to predict 1-year recurrence and recurrence-free survival in patients after CRLM resection. In light of the complexity and redundancy of features in the liver region, we designed the multi-modal guided local feature fusion module to utilize the tumor features to guide the dynamic fusion of prognostically relevant local features within the liver. On the other hand, to solve the loss of spatial information during multi-sequence MRI fusion, the cross-modal complementary external attention module designed an external mask branch to establish inter-layer correlation. The results show that the model has accuracy (ACC) of 0.79, the area under the curve (AUC) of 0.84, C-Index of 0.73, and hazard ratio (HR) of 4.0, which is a significant improvement over state-of-the-art methods. Additionally, MGCNet exhibits good interpretability.

Brain MRI and CT Image Fusion Using Multiscale Local Extrema and Image Statistics

In medical applications such as radiotherapy and guided-image surgery, data fusion for diagnostic imaging has emerged as a critical issue. Because the objective of medical image fusion is to improve patient diagnosis accuracy, the fused image is created by preserving the source images' prominent details and features. It has been demonstrated that the Multi-Level Local Extrema representation has numerous advantages over traditional image modeling approaches. We propose an innovative MLE-based fusion method for multimodal medical images in this paper. In the MLE schema, the proposed algorithm decomposes source images into coarse and detailed layers, then fuses the source images using weights calculated from these detail images using image statistics. We visually and quantitatively compared the efficacy of the suggested approach to that of existing methods using five different types of medical images from various sources. The experimental results showed that the proposed scheme outperforms other current typical schemes in terms of both qualitative image quality and objective evaluation.

Attention-Enhanced Fusion of Structural and Functional MRI for Analyzing HIV-Associated Asymptomatic Neurocognitive Impairment.

Asymptomatic neurocognitive impairment (ANI) is a predominant form of cognitive impairment among individuals infected with human immunodeficiency virus (HIV). The current diagnostic criteria for ANI primarily rely on subjective clinical assessments, possibly leading to different interpretations among clinicians. Some recent studies leverage structural or functional MRI containing objective biomarkers for ANI analysis, offering clinicians companion diagnostic tools. However, they mainly utilize a single imaging modality, neglecting complementary information provided by structural and functional MRI. To this end, we propose an attention-enhanced structural and functional MRI fusion (ASFF) framework for HIV-associated ANI analysis. Specifically, the ASFF first extracts data-driven and human-engineered features from structural MRI, and also captures functional MRI features via a graph isomorphism network and Transformer. A mutual cross-attention fusion module is then designed to model the underlying relationship between structural and functional MRI. Additionally, a semantic inter-modality constraint is introduced to encourage consistency of multimodal features, facilitating effective feature fusion. Experimental results on 137 subjects from an HIV-associated ANI dataset with T1-weighted MRI and resting-state functional MRI show the effectiveness of our ASFF in ANI identification. Furthermore, our method can identify both modality-shared and modality-specific brain regions, which may advance our understanding of the structural and functional pathology underlying ANI.

The role of PSA kinetics in men with a negative MRI-targeted prostate biopsy.

To evaluate rebiopsy rates and clinicopathologic outcomes in patients after a negative MRI-guided biopsy to better inform the management of these patients. Patients were included with a clinical suspicion of prostate cancer (PCa) referred for fusion biopsy for a PI-RADS v2.1 lesion ≥ 3 on multiparametric MRI and a negative MRI fusion biopsy. Biopsies included targeted and systematic cores. Patients with a prior cancer diagnosis were excluded. Both baseline and follow-up clinicopathological data, and long-term PSA values were examined in these patients. Statistical analyses included Wilcoxon rank-sum test and one-way tests. Of 685 total patients, 188 (27%) had a negative fusion biopsy. Of these 88 (47%), 74 (39%), and 26 (14%) had PI-RADS 3, 4, 5 lesions, respectively. Complete follow-up was available for 182/188 patients (97%), with a median of 24 months (interquartile range: 12-38). Post-biopsy PSA levels decreased the first and the second year (-0.24; and -0.84 ng/ml/yrs respectively). In follow-up, 44 patients had an MRI (24%) and 20 had a biopsy (10%). A positive PSA velocity was the only predictive variable for repeat MRI in univariate analysis. On repeat MRI, 9 (27%) patients had disappearance of the initial lesion, 21 (48%) had a lower PIRADS score and 14 (32%) higher. Only 12/182 (6.6%) were found to have PCa during follow-up, of those 7 (3.8%) were clinically significant. For patients with nonmalignant biopsy findings after an initial mpMRI showing a suspicious PI-RADS lesion, the majority of patients will have their PSAs return to baseline over time. To support this, repeat MRI frequently demonstrated a disappearance or downgrading of PIRADS lesions. These data support monitoring patients with this clinical scenario.

Comparative diagnostic accuracy of multiparametric magnetic resonance imaging-ultrasound fusion-guided biopsy versus systematic biopsy for clinically significant prostate cancer.

To examine the accuracy of transperineal magnetic resonance imaging (MRI)-ultrasound (US) fusion biopsy (FB) in identifying men with prostate cancer (PCa) that has reached a clinically relevant stage. This investigation enrolled 459 males. In 210 of these patients (FB group), transperineal MRI/US fusion-guided biopsies were performed on the suspicious region, and in 249 others, a systematic biopsy (SB) was performed (SB group). We compared these groups using Gleason scores and rates of cancer detection. PCa cases counted 198/459 (43.1%), including 94/249 (37.8%) in the SB group and 104/210 (49.5%) in the FB group. FB was associated with higher overall diagnostic accuracy relative to SB (88.5% and 72.3%, P = 0.024). FB exhibited greater sensitivity than SB (88.9% and 71.2%, P = 0.025). The area under the curve for FB and SB approaches was 0.837 and 0.737, respectively, such that FB was associated with an 11.9% increase in accuracy as determined based upon these AUC values. Relative to SB, FB was better able to detect high-grade tumors (GS ≥ 7) (78.85% vs. 60.64%, P = 0.025). Transperineal MRI-US fusion targeted biopsy is superior to the systematic one as an approach to diagnosing clinically significant PCa, as it is a viable technical approach to prostate biopsy.

A few-shot learning-based ischemic stroke segmentation system using weighted MRI fusion

Stroke, particularly ischemic stroke, is a major cause of disability and one of the leading causes of adult mortality worldwide. Early and prompt management of stroke patients can reduce the severity of the disease. Doctors usually determine the severity of a stroke by focusing on the region of interest (ROI) in the MRI or CT scan images. An accurate and effective automatic image segmentation system can assist medical professionals as well as automatic detection and classification systems. Deep learning is the current advanced approach for dealing with machine learning and artificial intelligence. However, conventional deep learning requires a large amount of data for training, and the amount of labeled data in the medical field is limited. In this paper, we propose a few-shot learning strategy and integrate it with a base convolutional neural network model, which utilizes a self-attention mechanism to segment MRI for ischemic stroke. By combining the base model with self-attention, we can focus more on the ROI and disregard less important features. Additionally, the proposed system only selects slices with lesions and ignores unlesioned slices. This helps to improve efficiency and reduce the computational load by eliminating the need to tune unnecessary parameters. To achieve even better results, the system also combines two weighted images, FLAIR and DWI, as an early fusion process. Experiments have shown that this approach leads to higher performance compared to using the same system without fusion. The proposed system is evaluated using a publicly available dataset, ISLES 2015 SSIS, and compared with other state-of-the-art (SOTA) systems. It achieves a dice coefficient score of 0.68, which is significantly better than that of other SOTA systems.

MRI and CT Fusion in Stereotactic Electroencephalography (SEEG).

Epilepsy is a neurological disorder characterized by spontaneous recurrent seizures. While 20% to 30% of epilepsy cases are untreatable with Anti-Epileptic Drugs, some of these cases can be addressed through surgical intervention. The success of such interventions greatly depends on accurately locating the epileptogenic tissue, a task achieved using diagnostic techniques like Stereotactic Electroencephalography (SEEG). SEEG utilizes multi-modal fusion to aid in electrode localization, using pre-surgical resonance and post-surgical computer tomography images as inputs. To ensure the absence of artifacts or misregistrations in the resultant images, a fusion method that accounts for electrode presence is required. We proposed an image fusion method in SEEG that incorporates electrode segmentation from computed tomography as a sampling mask during registration to address the fusion problem in SEEG. The method was validated using eight image pairs from the Retrospective Image Registration Evaluation Project (RIRE). After establishing a reference registration for the MRI and identifying eight points, we assessed the method's efficacy by comparing the Euclidean distances between these reference points and those derived using registration with a sampling mask. The results showed that the proposed method yielded a similar average error to the registration without a sampling mask, but reduced the dispersion of the error, with a standard deviation of 0.86 when a mask was used and 5.25 when no mask was used.

Feasibility and Safety of Single-Isocenter/Multi-Lesion (SIML) HyperArc Brain SRT: Clinical Implementation and Early Outcomes

Patients with multiple brain metastases may not tolerate relatively longer treatment times for traditional stereotactic radiation therapy (SRT) with individual isocenter plans for each lesion due to discomfort or co-morbidities. SRT using a single-isocenter/multi-lesion (SIML) HyperArc volumetric modulated arc therapy (VMAT) plan with flattening filter free (FFF) beam could significantly shorten overall treatment time, and improve patient comfort, compliance, and clinic efficiency. We report early clinical results of treating multiple brain metastases with SIML HyperArc SRT. Twenty-three patients with multiple brain metastatic tumors (range, 2-9 lesions; total treated lesions, n = 96) were simulated using Encompass support and Q-fix mask, and treated with a highly-conformal SIML VMAT SRT plans via non-coplanar HyperArc geometry. Mean tumor distance to isocenter was 5.3 cm, maximum up to 7 cm. Common prescriptions were 25-30 Gy/5 fractions, 24-27 Gy/3 fractions, and 20 Gy/1 fraction prescribed to each planning target volume (PTV) using 2 mm margin around standard gross tumor volume (GTV) delineated on contrasted enhanced MP-RAGE MRI fusion. Acuros dose calculation for 6MV-FFF beam was used for tissue heterogeneity corrections. Alliance A071801 criteria was used for dose constraints to organs at risk (OAR) and target conformality. Treatment was delivered every other day with CBCT-guidance, adjustments made with 6DOF couch corrections on a medical linear accelerator, and treatment delivery time within 15 minutes. Local control rates were reported, and toxicity profile rated based on CTCAE v5.0 for brain radionecrosis, optic neuropathy, and brainstem dysfunction. All plans met Alliance A071801 requirements for each tumor coverage, dose to OAR including optic apparatus, brainstem, and spinal cord. Mean GTV and PTV volume were 9.4 cc (range, 0.3-54.8 cc) and 16.13 cc (range, 1.0-80.2 cc). Patient-specific quality assurance results were 98.3% for gamma passing criteria of 2%/2mm. Independent in-house Monte Carlo physics second check agreed with HyperArc plans by ±3.0%. Mean follow up was 6 months (range, 0.0-18.6 months). Of the 23 patients treated, 17 (74%) had post-treatment MRI imaging to assess local control and toxicity. Local control was achieved in 69/73 (95%) of treated and followed lesions. CTCAE grade 2 radionecrosis occurred in 2 patients and were managed with dexamethasone. No CTCAE grade 3+ events of radionecrosis, optic pathway dysfunction, or brainstem toxicity were observed. SIML HyperArc Brain SRT for multiple brain metastases has excellent local control and low toxicity profile in our patients. It can significantly reduce treatment delivery time as compared to traditional multiple-isocenter brain SRT or chronologically separate treatment courses and thus, help to improve patient comfort, compliance, ease of care, and clinic workflow. Longer median follow up of SIML brain SRT on larger patient cohort is warranted.