Manual Segmentation Research Articles

Exploring the Impact of Variability in Cell Segmentation and Tracking Approaches.

Segmentation and tracking are essential preliminary steps in the analysis of almost all live cell imaging applications. Although the number of open-source software systems that facilitate automated segmentation and tracking continue to evolve, many researchers continue to opt for manual alternatives for samples that are not easily auto-segmented, tracing cell boundaries by hand and reidentifying cells on consecutive frames by eye. Such methods are subject to inter-user variability, introducing idiosyncrasies into the results of downstream analysis that are a result of subjectivity and individual expertise. The methods are also susceptible to intra-user variability, meaning findings are challenging to reproduce. In this pilot study, we demonstrate and quantify the degree of intra- and inter-user variability in manual cell segmentation and tracking by comparing the phenotypic metrics extracted from cells segmented and tracked by different members of our research team. Furthermore, we compare the segmentation results for a ptychographic cell image obtained using different automated software and demonstrate the high dependence of performance on the imaging modality they were developed to handle. Our results show that choice of segmentation and tracking methods should be considered carefully in order to enhance the quality and reproducibility of results.

Magnetic Resonance Imaging Texture Analysis Based on Intraosseous and Extraosseous Lesions to Predict Prognosis in Patients with Osteosarcoma

Objectives: To construct an optimal magnetic resonance imaging (MRI) texture model to evaluate histological patterns and predict prognosis in patients with osteosarcoma (OS). Methods: Thirty-four patients underwent pretreatment MRI and were diagnosed as having OS by surgical resection or biopsy between September 2008 and June 2018. Histological patterns and 3-year survival were recorded. Manual segmentation was performed in intraosseous, extraosseous, and entire lesions on T1-weighted, T2-weighted, and contrast-enhanced T1-weighted images to extract texture features and perform principal component analysis. A support vector machine algorithm with 3-fold cross-validation was used to construct and validate the models. The area under the receiver operating characteristic curve (AUC) was calculated to evaluate diagnostic performance in evaluating histological patterns and 3-year survival. Results: Eight patients were chondroblastic and the remaining twenty-six patients were non-chondroblastic patterns. Twenty-seven patients were 3-year survivors, and the remaining seven patients were non-survivors. In discriminating chondroblastic from non-chondroblastic patterns, the model from extraosseous lesions on the T2-weighted images showed the highest diagnostic performance (AUCs of 0.94 and 0.89 in the training and validation sets). The model from intraosseous lesions on the T1-weighted images showed the highest diagnostic performance in discriminating 3-year non-survivors from survivors (AUCs of 0.99 and 0.88 in the training and validation sets) with a sensitivity, specificity, positive predictive value, and negative predictive value of 85.7%, 92.6%, 75.0%, and 96.2%, respectively. Conclusions: The texture models of extraosseous lesions on T2-weighted images can discriminate the chondroblastic pattern from non-chondroblastic patterns, while the texture models of intraosseous lesions on T1-weighted images can discriminate 3-year non-survivors from survivors.

Application of deep learning algorithms for identifying deterioration in the ushnisha (Head Bun) of the Leshan Giant Buddha

AbstractThe Leshan Giant Buddha’s ushnisha (Head Bun) has suffered from loss of lime plaster, cracks, and biological damage, compromising its structural integrity and reducing the effectiveness of the drainage system in the Buddha's head. The infiltration of moisture has led to water damage within the statue, significantly accelerating its weathering. This situation urgently requires protection and reinforcement measures. Detecting deterioration in the ushnisha is a crucial step in the preservation process. In this study, we utilized two deep learning models for pixel-level semantic segmentation of the damage. Due to the small size of the cracks, a weighted loss function was applied to improve both the training speed of the model and the efficiency of crack identification. This weighting strategy proved effective for both models. The weighted K-Net model achieved a mean accuracy (mAcc) of 90.23% and a mean intersection-over-union (mIoU) of 69.55%, with a damage segmentation speed of 7 images per second, which is 1309 times faster than manual segmentation. By applying the trained deep learning models to re-examine the ushnisha, we successfully identified damage that had been overlooked during manual annotation. Using the model’s enhanced results, we conducted a comprehensive quantification of the damage across all ushnisha and identified the most severely affected areas. Additionally, we performed a model interpretability analysis to explain the decision-making process and principles of the deep learning models. This research provides significant practical value for detecting and quantifying damage in the Leshan Giant Buddha.

A triplanar 2D convolutional network ensemble for infra-alveolar nerve segmentation

AbstractCone Beam Computed Tomography scan segmentation is a key step in the digital dentistry workflow. This technology is considered the gold standard for imaging in implant dentistry as well as maxillofacial and oral surgery, where extracting volumetric information is essential. However, manual segmentation is a tedious process that requires significant time and expertise. Therefore, to circumvent these issues, we have developed and present in this study a novel tool for automated nerve segmentation based on deep learning. Specifically, we build on nnU-Net’s self-configuring framework by constructing an ensemble of three 2D convolutional models that leverage information extracted from the three orthogonal planes: axial, coronal and sagittal. We tested our method by participating in the Challenge from the $$26^{th}$$ 26 th edition of the International Conference on Medical Image Computing and Computer Assisted Interventions, held in 2023, where we achieved a Dice Similarity Coefficient of 0.66 and a $$95^{th}$$ 95 th percentile Hausdorff distance of 8.63.

INNV-25. POST-OPERATIVE LOW DOSE VERSUS STANDARD DOSE DEXAMETHASONE FOR MANAGEMENT OF CEREBRAL EDEMA: A RANDOMIZED, DOUBLE-BLINDED FEASIBILITY STUDY IN PATIENTS WITH PRIMARY OR METASTATIC BRAIN TUMORS UNDERGOING CRANIOTOMY FOR TUMOR RESECTION

Abstract INTRODUCTION Although dexamethasone is effective for controlling cerebral edema during the peri-operative period, its immunosuppressive properties can inhibit responses to immunotherapy. We hypothesized that patients with mild to moderate cerebral edema could be safely managed post-operatively with lower doses of dexamethasone. METHODS Eligible patients with brain tumors who had less than 10 mm of midline shift on their pre-operative brain MRIs were randomized 1:1 to the standard or low dose dexamethasone arm, each tapering double blinded through post-op day 8. The double blind was maintained using equivalent volumes for IV doses and capsules containing dexamethasone or placebo. By post-op day 9, participants received the same dose of dexamethasone and continued tapering to off by day 14. Additional doses of dexamethasone were allowed if needed, based on specified symptom criteria. Lack of feasibility was defined as a participant requiring 18 mg or more of dexamethasone in addition to the planned total dose of dexamethasone (low dose: 49 mg; standard dose: 112 mg). Brain MRIs were performed on post-operative days 1, 3, and 8 to evaluate changes in cerebral edema. Manual segmentation and computer-assisted techniques were used to quantitatively measure cerebral edema volume. RESULTS Twenty-six patients were randomized and treated within tumor type strata (primary versus metastatic), of which 25 were evaluable. None of the participants on the low dose dexamethasone arm met the criteria for lack of feasibility. The quantitative assessment of differences in cerebral edema between the arms is ongoing and will be presented. CONCLUSIONS Administration of lower doses of dexamethasone post-operatively to brain tumor patients with mild to moderate cerebral edema is feasible. Treatment with lower doses of dexamethasone post-operatively should be considered in this patient population when peri-operative administration of immunotherapy is planned.

CTNI-59. TUMOUR VOLUMES PREDICT PROGNOSIS AND RESPONSE TO TREATMENT WITH DEPATUXIZUMAB MAFADOTIN

Abstract BACKGROUND Preclinical data demonstrates that tumor volumes predicted response to the anti-EGFR ADC Depatuxizumab mafadotin (Depatux-M) because of reduced drug delivery in large tumors. We investigated the impact of tumor volumes on outcomes with Depatux-M in the EORTC randomized phase 2 Intellance 2 study on recurrent glioblastoma (NeuroOncol 22(5):684). METHODS 260 patients were randomized to Depatux-M, Depatux-M with temozolomide, or control treatment (lomustine/temozolomide). Manual segmentation of enhancing tumor on baseline MRI scans was performed and compared with OS and PFS using SPSS Version 29.0: Kaplan-Meier survival analysis including log rank test and Cox regression analyses; categorical data were compared with Fishers’ exact test; correlation between 2D and 3D measurements with Pearson’s correlation. RESULTS MRI scans and data from 240/260 (92%) patients were available. Firstly, we confirmed that tumor volumes were a significant prognostic variable. Significant differences were seen in OS between the 0-20 vs 20-40 ml groups (321 vs 216 days, p<0.001) and between the 20-40ml vs 40+ groups (216 vs 150 days, p<0.001) with corresponding differences in PFS. Tumor volumes were also predictive of benefit from Depatux-M, with only patients with smaller tumors benefiting from Depatux-M. Patients in the 0-20 ml and 20-40 ml group group had superior OS when treated with Depatux-M compared to controls (356 vs 309 days, p=0.011 and 229 vs 178 days, p=0.028 respectively). Conversely, the 40 ml + group had no benefit (146 vs 152 days, p=0.875) as predicted. The benefit was maximal in the combination arm, with OS improvement in the 0-20 ml group (437 vs 309 days, p=0.007) and 20-40ml groups (225 vs 178 days, p=0.009). The impact of tumor volumes was confirmed in multi-variate analysis and superior to similar analysis with 2D RANO measurements. CONCLUSIONS Tumor volumes are an important prognostic variable in recurrent glioblastoma and modulate survival to large molecules like Depatux-M.

Enhancing the diagnostic capacity of [18F]PSMA-1007 PET/MRI in primary prostate cancer staging with artificial intelligence and semi-quantitative DCE: an exploratory study

BackgroundTo investigate the ability of artificial intelligence (AI)-based and semi-quantitative dynamic contrast enhanced (DCE) multiparametric MRI (mpMRI), performed within [18F]-PSMA-1007 PET/MRI, in differentiating benign from malignant prostate tissues in patients with primary prostate cancer (PC).ResultsA total of seven patients underwent whole-body [18F]-PSMA-1007 PET/MRI examinations including a pelvic mpMRI protocol with T2w, diffusion weighted imaging (DWI) and DCE image series. Conventional analysis included visual reading of PET/MRI images and Prostate Imaging Reporting & Data System (PI-RADS) scoring of the prostate. On the prostate level, we performed manual segmentations for time-intensity curve parameter formation and semi-quantitative analysis based on DCE segmentation data of PC-suspicious lesions. Moreover, we applied a recently introduced deep learning (DL) pipeline previously trained on 1010 independent MRI examinations with systematic biopsy-enhanced histopathological targeted biopsy lesion ground truth in order to perform AI-based lesion detection, prostate segmentation and derivation of a deep learning PI-RADS score. DICE coefficients between manual and automatic DL-acquired segmentations were compared. On patient-based analysis, PET/MRI revealed PC-suspicious lesions in the prostate gland in 6/7 patients (Gleason Score-GS ≥ 7b) that were histologically confirmed. Four of these patients also showed lymph node metastases, while two of them had bone metastases. One patient with GS 6 showed no PC-suspicious lesions. Based on DCE segmentations, a distinction between PC-suspicious and normal appearing tissue was feasible with the parameters fitted maximum contrast ratio (FMCR) and wash-in-slope. DICE coefficients (manual vs. deep learning) were comparable with literature values at a mean of 0.44. Further, the DL pipeline could identify the intraprostatic PC-suspicious lesions in all six patients with clinically significant PC.ConclusionFirstly, semi-quantitative DCE analysis based on manual segmentations of time-intensity curves was able to distinguish benign from malignant tissues. Moreover, DL analysis of the MRI data could detect clinically significant PC in all cases, demonstrating the feasibility of AI-supported approaches in increasing diagnostic certainty of PSMA-radioligand PET/MRI.

Quantitative modeling of lenticulostriate arteries on 7-T TOF-MRA for cerebral small vessel disease.

We developed a framework for segmenting and modeling lenticulostriate arteries (LSAs) on 7-T time-of-flight magnetic resonance angiography and tested its performance on cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) patients and controls. We prospectively included 29 CADASIL patients and 21 controls. The framework includes a small-patch convolutional neural network (SP-CNN) for fine segmentation, a random forest for modeling LSAs, and a screening model for removing wrong branches. The segmentation performance of our SP-CNN was compared to competitive networks. External validation with different resolution was performed on ten patients with aneurysms. Dice similarity coefficient (DSC) and Hausdorff distance (HD) between each network and manual segmentation were calculated. The modeling results of the centerlines, diameters, and lengths of LSAs were compared against manual labeling by four neurologists. The SP-CNN achieved higher DSC (92.741 ± 2.789, mean ± standard deviation) and lower HD (0.610 ± 0.141 mm) in the segmentation of LSAs. It also outperformed competitive networks in the external validation (DSC 82.6 ± 5.5, HD 0.829 ± 0.143 mm). The framework versus manual difference was lower than the manual inter-observer difference for the vessel length of primary branches (median -0.040 mm, interquartile range -0.209 to 0.059 mm) and secondary branches (0.202 mm, 0.016-0.537 mm), as well as for the offset of centerlines of primary branches (0.071 mm, 0.065-0.078 mm) and secondary branches (0.072, 0.064-0.080 mm), with p < 0.001 for all comparisons. Our framework for LSAs modeling/quantification demonstrated high reliability and accuracy when compared to manual labeling. NCT05902039 ( https://clinicaltrials.gov/study/NCT05902039?cond=NCT05902039 ). The proposed automatic segmentation and modeling framework offers precise quantification of the morphological parameters of lenticulostriate arteries. This innovative technology streamlines diagnosis and research of cerebral small vessel disease, eliminating the burden of manual labeling, facilitating cohort studies and clinical diagnosis. The morphology of LSAs is important in the diagnosis of CSVD but difficult to quantify. The proposed algorithm achieved the performance equivalent to manual labeling by neurologists. Our method can provide standardized quantitative results, reducing radiologists' workload in cohort studies.

Quantification of pulmonary edema using automated lung segmentation on computed tomography in mechanically ventilated patients with acute respiratory distress syndrome

BackgroundQuantification of pulmonary edema in patients with acute respiratory distress syndrome (ARDS) by chest computed tomography (CT) scan has not been validated in routine diagnostics due to its complexity and time-consuming nature. Therefore, the single-indicator transpulmonary thermodilution (TPTD) technique to measure extravascular lung water (EVLW) has been used in the clinical setting. Advances in artificial intelligence (AI) have now enabled CT images of inhomogeneous lungs to be segmented automatically by an intensive care physician with no prior radiology training within a relatively short time. Nevertheless, there is a paucity of data validating the quantification of pulmonary edema using automated lung segmentation on CT compared with TPTD.MethodsA retrospective study (January 2016 to December 2021) analyzed patients with ARDS, admitted to the intensive care unit of the Department of Anesthesiology and Critical Care Medicine, University Hospital Mannheim, who underwent a chest CT scan and hemodynamic monitoring using TPTD at the same time. Pulmonary edema was estimated using manually and automated lung segmentation on CT and then compared to the pulmonary edema calculated from EVLW determined using TPTD.Results145 comparative measurements of pulmonary edema with TPTD and CT were included in the study. Estimating pulmonary edema using either automated lung segmentation on CT or TPTD showed a low bias overall (− 104 ml) but wide levels of agreement (upper: 936 ml, lower: − 1144 ml). In 13% of the analyzed CT scans, the agreement between the segmentation of the AI algorithm and a dedicated investigator was poor. Manual segmentation and automated segmentation adjusted for contrast agent did not improve the agreement levels.ConclusionsAutomated lung segmentation on CT can be considered an unbiased but imprecise measurement of pulmonary edema in mechanically ventilated patients with ARDS.

Evaluation of registration-based vs. manual segmentation of rhesus macaque brain MRIs.

With increasing numbers of magnetic resonance imaging (MRI) datasets becoming publicly available, researchers and clinicians alike have turned to automated methods of segmentation to enable population-level analyses of these data. Although prior research has evaluated the extent to which automated methods recapitulate "gold standard" manual segmentation methods in the human brain, such an evaluation has not yet been carried out for segmentation of MRIs of the macaque brain. Macaques offer the important opportunity to bridge gaps between microanatomical studies using invasive methods like tract tracing, neural recordings, and high-resolution histology and non-invasive macroanatomical studies using methods like MRI. As such, it is important to evaluate whether automated tools derive data of sufficient quality from macaque MRIs to bridge these gaps. We tested the relationship between automated registration-based segmentation using an open source and actively maintained NHP imaging analysis pipeline (AFNI) and gold standard manual segmentation of 4 structures (2 cortical: anterior cingulate cortex and insula; 2 subcortical: amygdala and caudate) across 37 rhesus macaques (Macaca mulatta). We identified some variability in the strength of correlation between automated and manual segmentations across neural regions and differences in relationships with demographic variables like age and sex between the two techniques.

Automated dentition segmentation: 3D UNet-based approach with MIScnn framework

Advancements in technology have led to the adoption of digital workflows in dentistry, which require the segmentation of regions of interest from cone-beam computed tomography (CBCT) scans. These segmentations assist in diagnosis, treatment planning, and research. However, manual segmentation is an expensive and labor-intensive process. Therefore, automated methods, such as convolutional neural networks (CNNs), provide a more efficient way to generate segmentations from CBCT scans. A three-dimensional UNet-based CNN model, utilizing the Medical Image Segmentation CNN framework, was used for training and generating predictions from CBCT scans. A dataset of 351 CBCT scans, with ground-truth labels created through manual segmentation using AI-assisted segmentation software, was prepared. Data preprocessing, augmentation, and model training were performed, and the performance of the proposed CNN model was analyzed. The CNN model achieved high accuracy in segmenting maxillary and mandibular teeth from CBCT scans, with average Dice Similarity Coefficient values of 91.83% and 91.35% for maxillary and mandibular teeth, respectively. Performance metrics, including Intersection over Union, precision, and recall, further confirmed the model's effectiveness. The study demonstrates the efficacy of the three-dimensional UNet-based CNN model within the Medical Image Segmentation CNN framework for automated segmentation of maxillary and mandibular dentition from CBCT scans. Automated segmentation using CNNs has the potential to deliver accurate and efficient results, offering a significant advantage over traditional segmentation methods.

Impact of Artificial Intelligence-Based Autosegmentation of Organs at Risk in Low- and Middle-Income Countries

Radiation therapy (RT) processes require significant human resources and expertise, creating a barrier to rapid RT deployment in low- and middle-income countries (LMICs). Accurate segmentation of tumor targets and organs at risk (OARs) is crucial for optimal RT. This study assessed the impact of artificial intelligence (AI)-based autosegmentation of OARs in 2 LMICs. Ten patients, comprising 5 head and neck (HN) cancer patients and 5 prostate cancer patients, were randomly selected. Planning computed tomography images were subjected to autosegmentation using an Food and Drug Administration-approved AI software tool and manual segmentation by experienced radiation oncologists from 2 LMIC RT clinics. The control data, obtained from a large academic institution in the United States, consisted of contours obtained by an experienced radiation oncologist. The segmentation time, DICE similarity coefficient (DSC), Hausdorff distance, and mean surface distance were evaluated. AI significantly reduced segmentation time, averaging 2 minutes per patient, compared with 57 to 84 minutes for manual contouring in LMICs. Compared with the control data, the AI pelvic contours provided better agreement than did the LMIC manual contours (mean DSC of 0.834 vs 0.807 in LMIC1 and 0.844 vs 0.801 in LMIC2). For HN contours, AI provided better agreement for the majority of OAR contours than manual contours in LMIC1 (mean DSC: 0.823 vs 0.821) or LMIC2 (mean DSC: 0.792 vs 0.748). Neither the AI nor LMIC manual contours had good agreement with the control data (DSC < 0.600) for the optic nerves, chiasm, and cochlea. AI-based autosegmentation generates OAR contours of comparable quality to manual segmentation for both pelvic and HN cancer patients in LMICs, with substantial time savings.

Anatomical Analysis of Vertebral Arteries in Vertebrobasilar Dolichoectasia: a Multi-center Study

BackgroundVertebrobasilar dolichoectasia (VBD) is a rare disease with significant morbidity. Its propensity for posterior circulation and relationship with aneurysms is poorly understood. Here, we aimed to describe the anatomical characteristics of the vertebral arteries (VA) in patients with VBD. MethodsWe conducted a multi-center retrospective cohort analysis on patients diagnosed with VBD between January 2009 and December 2022. Anatomical variables were measured after manual segmentation on CTA. Length of the VAs, basilar artery (BA), sectional diameters, and tortuosity index were measured and compared with a control group. Results124 patients were included: 38 patients with VBD and 86 controls. VBD patients had longer VAs (right VA 262.5 vs. 228.7mm, p<0.01; left VA 253.4 vs. 222.1mm, p<0.01), longer BAs (41.9 vs. 30.0mm, p<0.01), and larger cross-sectional diameters of right V3 (5.1 vs. 4.6mm, p<0.05) and V4 segments (4.5 vs.2.9mm, p<0.05) and left V1-V4 segments (V1 6.1 vs. 4.2, V2 7.2 vs.4.8mm, V3 6.4 vs.5.0mm, V4 6.5 vs. 3.0mm, p<0.01). Patients with VBD and fusiform aneurysms had longer VAs and BA than patients without aneurysms (p<0.01). The VAs tortuosity index was higher in VBD than controls (60.69 vs. 44.18, p<0.01). No cases of vertebral stenosis were detected. ConclusionsOur study offers an anatomical study of VBD. While the natural history is poorly understood, the higher tortuosity index with associated wall shear stress, provides a possible mechanism for disease progression. Further studies are needed to better understand the

Shape modelling reveals age-related knee bony shape changes in asymptomatic knees.

Osteoarthritis (OA) causes bony shape changes within the knee. Furthermore, the risk of developing OA increases with age. However, age alone does not cause OA. It is therefore important to understand the healthy age-related trajectories of knee shape before attributing these changes to OA. The aim of this study was to determine the association between bony knee shape and age using statistical-shape modelling (SSM). 96 participants received a CT scan of their knee. Three-dimensional models were created using manual segmentation. Separate SSM's for the distal femur and proximal tibia were created. Linear regression models were used to assess the association between age and femoral and tibial shape. Fourteen modes of the femoral and tibial SSM's captured 68% and 73% shape variation, respectively. Only femoral mode 3 and tibial mode 7 were associated with age. Increasing age was related to larger femoral bone volume and deepening of the femoral trochlear groove. Furthermore, increased age was associated with medial tibial plateau expansion. Aspects of bony femoral and tibial shape were significantly associated with aging, including femoral and tibial bone size, femoral trochlear groove, and medial tibial plateau area. Changes in knee morphology occur as a normal process of aging without osteoarthritis development. This may be a response to mechanical loading over time. Further research investigating the effect of these changes on loading in the knee may provide valuable information for knee health in older age.

Automated Dental Registration and TMJ Segmentation for Virtual Surgical Planning of Orthognathic Surgery via Three-Step Computer-Based Method

Objective: This study developed and evaluated a computer-based method for automating the registration of scanned dental models with 3D reconstructed skulls and segmentation of the temporomandibular joint (TMJ). Methods: A dataset comprising 1274 skull models and corresponding scanned dental models was collected. In total, 1066 cases were used for the development of the computer-based method, while 208 cases were used for validation. Performance was evaluated by comparing the automated results with manual registration and segmentation performed by clinicians, using accuracy and completeness metrics (e.g. intersection of union [IoU] and Dice similarity coefficient [DSC]). Results: The automated registration achieved a mean absolute error of 0.35 mm for the maxilla and 0.38 mm for the mandible, and a root mean squared error of 0.46 mm and 0.39 mm, respectively. The automatic TMJ segmentation exhibited an accuracy of 97.48%, a precision of 97.06%, a IoU of 95.72%, DSC of 97.3%, and a Hausdorff value of 1.87 mm, which were sufficient for clinical application. Conclusion: The proposed method significantly improved the efficiency of orthognathic surgical planning by automating the registration and segmentation processes. The accuracy and precision of the automated results were sufficient for clinical use, reducing the workload on clinicians and facilitating faster and more reliable surgical planning. Clinical significance: The computer-based method streamlines orthognathic surgical planning, enhancing precision and efficiency without compromising clinical accuracy, ultimately improving patient outcomes and reducing the workload of surgeons.

Higher Posterior Tibial Slope and Lower Medial Proximal Tibial Angle of the Knee Are Associated with Delayed Graft Maturation after ACL Reconstruction Surgery with Hamstring Autograft

PurposeThis study aims to relate bone shape with anterior cruciate ligament (ACL) graft maturation, as evaluated by quantitative magnetic resonance imaging (MRI). MethodsWe retrospectively evaluated patients aged 18-60 years who underwent ACL reconstruction with a hamstring autograft with doubled semitendinosus and gracilis at our institution between 2018 and 2020 with isolated ACL injuries. All patients had a minimum follow-up period of 2 years. To evaluate alignment, radiographs of the knee were taken preoperatively. Postoperative 3T magnetic resonance imaging (MRI) and patient-reported outcomes (PRO) scores were acquired at 2 years after surgery. A combined T1ρ/T2 quantification sequence and a 3D multi-echo ultrashort echo-time (UTE) Cones sequence were obtained in a sagittal-oblique plane for better assessment of the ACL graft. After acquiring images, a manual segmentation for ACL graft was performed to assess T1ρ, T2 and UTE T2* relaxation times. Correlations between radiographic parameters or PRO scores and T1ρ, T2 and UTE T2* values were tested with Pearson correlation coefficients. Significance was defined as p<0.05. ResultsSubjects were 28 patients (12 males and 16 females) and had an average age of 34.3 years (SD=8.5 years) and a BMI of 24.4 kg/m2 (SD=4.5 kg/m2). A significant correlation was found between the preoperative PTS and the T1ρ, T2 and UTE T2* relaxation times (r=0.41; p=0.033, r =0.47; p=0.014 and r =0.41; p=0.030), indicating inferior graft maturation in patients with higher PTS. Furthermore, a negative correlation was identified between the preoperative MPTA and the T1ρ, T2 and UTE T2* relaxation times (r=-0.39; p=0.039, r =-0.46; p=0.018 and r =-0.43; p=0.024), indicating inferior graft maturation in patients with greater varus knee alignment. ConclusionThe study findings suggest that an increased PTS and lower MPTA are associated with inferior graft maturation as per T1ρ, T2 and UTE T2* imaging at 2 years post-ACL reconstruction. Level of EvidenceLevel IV retrospective case series

Prediction of femoral head collapse in osteonecrosis using deep learning segmentation and radiomics texture analysis of MRI

BackgroundFemoral head collapse is a critical pathological change and is regarded as turning point in disease progression in osteonecrosis of the femoral head (ONFH). In this study, we aim to build an automatic femoral head collapse prediction pipeline for ONFH based on magnetic resonance imaging (MRI) radiomics.MethodsIn the segmentation model development dataset, T1-weighted MRI of 222 hips from two hospitals were retrospectively collected and randomly split into training (n = 190) and test (n = 32) sets. In the prognosis prediction model development dataset, 206 hips were also retrospectively collected from two hospitals and divided into training set (n = 155) and external test set (n = 51) according to data source. A deep learning model for automatic lesion segmentation was trained with nnU-Net, from which three-dimensional regions of interest were segmented and a total of 107 radiomics features were extracted. After intra-class correlation coefficients screening, feature correlation coefficient screening and Least Absolute Shrinkage and Selection Operator regression feature selection, a machine learning model for ONFH prognosis prediction was trained with Logistic Regression (LR) and Light Gradient Boosting Machine (LightGBM) algorithm.ResultsThe segmentation model achieved an average dice similarity coefficient of 0.848 and an average 95% Hausdorff distance of 3.794 in the test set, compared to the manual segmentation results. After feature selection, nine radiomics features were included in the prognosis prediction model. External test showed that the LightGBM model exhibited acceptable predictive performance. The area under the curve (AUC) of the prediction model was 0.851 (95% CI: 0.7268–0.9752), with an accuracy of 0.765, sensitivity of 0.833, and specificity of 0.727. Decision curve analysis showed that the LightGBM model exhibited favorable clinical utility.ConclusionThis study presents an automated pipeline for predicting femoral head collapse in ONFH with acceptable performance. Further research is necessary to determine the clinical applicability of this radiomics-based approach and to assess its potential to assist in treatment decision-making for ONFH.

Automatic tumor segmentation and lymph node metastasis prediction in papillary thyroid carcinoma using ultrasound keyframes.

Accurate preoperative prediction of cervical lymph node metastasis (LNM) for papillary thyroid carcinoma (PTC) patients is essential for disease staging and individualized treatment planning, which can improve prognosis and facilitate better management. To establish a fully automated deep learning-enabled model (FADLM) for automated tumor segmentation and cervical LNM prediction in PTC using ultrasound (US) video keyframes. The bicentral study retrospective enrolled 518 PTC patients, who were then randomly divided into the training (Hospital 1, n=340), internal test (Hospital 1, n=83), and external test cohorts (Hospital 2, n=95). The FADLM integrated mask region-based convolutional neural network (Mask R-CNN) for automatic thyroid primary tumor segmentation and ResNet34 with Bayes strategy for cervical LNM diagnosis. A radiomics model (RM) using the same automated segmentation method, a traditional radiomics model (TRM) using manual segmentation, and a clinical-semantic model (CSM) were developed for comparison. The dice similarity coefficient (DSC) was used to evaluate segmentation performance. The prediction performance of the models was validated in terms of discrimination and clinical utility with the area under the receiver operator characteristic curve (AUC), heatmap analysis, and decision curve analysis (DCA). The comparison of the predictive performance among different models was conducted by DeLong test. The performances of two radiologists compared with FADLM and the diagnostic augmentation with FADLM's assistance were analyzed in terms of accuracy, sensitivity and specificity using McNemar's x2 test. The p-value less than 0.05 was defined as a statistically significant difference. The Benjamini-Hochberg procedure was applied for multiple comparisons to deal with Type I error. The FADLM yielded promising segmentation results in training (DSC: 0.88±0.23), internal test (DSC: 0.88±0.23), and external test cohorts (DSC: 0.85±0.24). The AUCs of FADLM for cervical LNM prediction were 0.78 (95% CI: 0.73, 0.83), 0.83 (95% CI: 0.74, 0.92), and 0.83 (95% CI: 0.75, 0.92), respectively. It all significantly outperformed the RM (AUCs: 0.78vs. 0.72; 0.83vs. 0.65; 0.83vs. 0.68, all adjusted p-values<0.05) and CSM (AUCs: 0.78vs. 0.71; 0.83vs. 0.62; 0.83vs. 0.68, all adjusted p-values<0.05) across the three cohorts. The RM offered similar performance to that of TRM (AUCs: 0.61vs. 0.63, adjusted p-value =0.60) while significantly reducing the segmentation time (3.3±3.8 vs. 14.1±4.2 s, p-value <0.001). Under the assistance of FADLM, the accuracies of junior and senior radiologists were improved by 18% and 15% (all adjusted p-values<0.05) and the sensitivities by 25% and 21% (all adjusted p-values<0.05) in the external test cohort. The FADLM with elaborately designed automated strategy using US video keyframes holds good potential to provide an efficient and consistent prediction of cervical LNM in PTC. The FADLM displays superior performance to RM, CSM, and radiologists with promising efficacy.

Automated segmentation of retinal layers in optical coherence tomography images using Xception70 feature extraction

Optical coherence tomography (OCT) imaging plays a critical role in evaluating retinal layer thickness, serving as a pivotal diagnostic tool for numerous retinal conditions. However, challenges such as speckle noise, poor image contrast, and ambiguous retinal detachments, like drusen, often hinder accurate segmentation of retinal layers. To address these challenges and enhance diagnostic precision, we introduce the Retinal Segmentation Network, or "Ret-Seg Net," a deep neural network-based approach. Leveraging the advanced Xception70 feature extractor, Ret-Seg Net extracts and comprehends the intricate properties of retinal layers. By integrating acquired feature maps from Xception70 into the atrous spatial pyramid-pooling module, Ret-Seg Net extracts multiscale feature information. The encoder-decoder module of Ret-Seg Net achieves automated segmentation of retinal layers in OCT images by reconstructing distinct retinal layer borders. This advanced module accurately recognizes and differentiates retinal layer boundaries, providing precise and reliable segmentation. Validation of our approach using real-time images and the Duke dataset, comprising 310 volumes with 40 B-scans each, demonstrates outstanding performance. Mean intersection over union (MIoU) and sensitivity (Se) metrics achieved remarkable values of 94.52 % and 96.25 % respectively. Furthermore, our approach offers a versatile segmentation framework applicable to various tissues and cell types in clinical settings. Automating segmentation of retinal layers enhances precision in disease identification and monitoring while significantly improving labor efficiency by reducing the need for manual segmentation.

A road map to manual segmentation of cerebral structures.

Manual segmentation is an essential tool in the researcher's technical arsenal. It is a frequent practice necessary for image analysis in many protocols, especially in neuroimaging and comparative brain anatomy. In the framework of emergence of studies focusing on alternative animal models, manual segmentation procedures play a critical role. Nevertheless, this critical task is often assigned to students, a process that, unfortunately, tends to be time-consuming and repetitive. Well-conducted and well-described segmentation procedures can potentially guide novice and even expert operators and enhance research works' internal and external validity, making it possible to harmonize studies and facilitate data sharing. Furthermore, recent advances in neuroimaging, such as exvivo imaging or ultra-high-field MRI, enable new acquisition modalities and the identification of minute structures that are barely visible with typical approaches. In this context of increasingly detailed and multimodal brain studies, reflecting on methodology is relevant and necessary. Because it is crucial to implement good practices in manual segmentation per se but also in the description of the segmentation procedures in research papers, we propose a general roadmap for optimizing the technique, its process and the reporting of manual segmentation. For each of them, the relevant elements of the literature have been collected and cited. The article is accompanied by a checklist that the reader can use to verify that the critical steps are being followed.