Detection Of Lesions Research Articles

Automated Detection of Oral Malignant Lesions Using Deep Learning: Scoping Review and Meta-Analysis.

Oral diseases, specifically malignant lesions, are serious global health concerns requiring early diagnosis for effective treatment. In recent years, deep learning (DL) has emerged as a powerful tool for the automated detection and classification of oral lesions. This research, by conducting a scoping review and meta-analysis, aims to provide an overview of the progress and achievements in the field of automated detection of oral lesions using DL. A scoping review was conducted to identify relevant studies published in the last 5 years (2018-2023). A comprehensive search was conducted using several electronic databases, including PubMed, Web of Science, and Scopus. Two reviewers independently assessed the studies for eligibility and extracted data using a standardized form, and a meta-analysis was conducted to synthesize the findings. Fourteen studies utilizing various DL algorithms were identified and included for the detection and classification of oral lesions from clinical images. Among these, three were included in the meta-analysis. The estimated pooled sensitivity and specificity were 0.86 (95% confidence interval [CI] = 0.80-0.91) and 0.67 (95% CI = 0.58-0.75), respectively. The results of meta-analysis indicate that DL algorithms improve the diagnosis of oral lesions. Future research should develop validated algorithms for automated diagnosis. Open Science Framework (https://osf.io/4n8sm).

AI-Assisted Detection and Localization of Spinal Metastatic Lesions.

The integration of machine learning and radiomics in medical imaging has significantly advanced diagnostic and prognostic capabilities in healthcare. This study focuses on developing and validating an artificial intelligence (AI) model using U-Net architectures for the accurate detection and segmentation of spinal metastases from computed tomography (CT) images, addressing both osteolytic and osteoblastic lesions. Our methodology employs multiple variations of the U-Net architecture and utilizes two distinct datasets: one consisting of 115 polytrauma patients for vertebra segmentation and another comprising 38 patients with documented spinal metastases for lesion detection. The model demonstrated strong performance in vertebra segmentation, achieving Dice Similarity Coefficient (DSC) values between 0.87 and 0.96. For metastasis segmentation, the model achieved a DSC of 0.71 and an F-beta score of 0.68 for lytic lesions but struggled with sclerotic lesions, obtaining a DSC of 0.61 and an F-beta score of 0.57, reflecting challenges in detecting dense, subtle bone alterations. Despite these limitations, the model successfully identified isolated metastatic lesions beyond the spine, such as in the sternum, indicating potential for broader skeletal metastasis detection. The study concludes that AI-based models can augment radiologists' capabilities by providing reliable second-opinion tools, though further refinements and diverse training data are needed for optimal performance, particularly for sclerotic lesion segmentation. The annotated CT dataset produced and shared in this research serves as a valuable resource for future advancements.

Image-Based Generative Artificial Intelligence in Radiology: Comprehensive Updates.

Generative artificial intelligence (AI) has been applied to images for image quality enhancement, domain transfer, and augmentation of training data for AI modeling in various medical fields. Image-generative AI can produce large amounts of unannotated imaging data, which facilitates multiple downstream deep-learning tasks. However, their evaluation methods and clinical utility have not been thoroughly reviewed. This article summarizes commonly used generative adversarial networks and diffusion models. In addition, it summarizes their utility in clinical tasks in the field of radiology, such as direct image utilization, lesion detection, segmentation, and diagnosis. This article aims to guide readers regarding radiology practice and research using image-generative AI by 1) reviewing basic theories of image-generative AI, 2) discussing the methods used to evaluate the generated images, 3) outlining the clinical and research utility of generated images, and 4) discussing the issue of hallucinations.

Effects of Gadolinium Chelate Administration Timing on T2-weighted and Diffusion-weighted Abdominal MRI Examination: A Prospective Study.

Magnetic Resonance Imaging (MRI) data acquisition includes several sequences that might be optimized to reduce the scan time. This study aimed to investigate the impact of gadolinium chelate administration timing on scan duration and image quality in Diffusion-weighted Imaging (DWI) and T2-weighted Imaging (T2WI) during abdominal MRI examinations. A prospective study was conducted from October 2018 to May 2020. Study participants were assigned into a conventional group, undergoing MRI with DWI and T2WI sequences pre and post-gadolinium injection, or an optimized group, receiving MRI with DWI and T2WI sequences after gadolinium injection. Quantitative image quality, measured by the Signal-to-noise Ratio (SNR), Contrast-to-noise Ratio (CNR), and Apparent Diffusion Coefficient (ADC), was analyzed. Kappa statistics were employed for the inter-observer agreement on liver lesion detection. Our study has included 341 patients, with 168 and 173 in the conventional and optimized groups, respectively. Mean scan durations were 1,304 (±143) and 1,015 (±129) s for the conventional and optimized groups, respectively (p<0.05). For the liver, spleen, and pancreas, SNR and ADC remained statistically unchanged in post-enhanced DWI and T2WI (p>0.05). Significant decreases in the SNR and ADC of the kidney were observed in post-contrast DWI and T2WI (p<0.05). Hepatic lesion detectability did not show significant differences between pre and post-contrast DWI and T2WI images (p>0.05). DWI and T2WI sequences assessed post-gadolinium administration exhibited shortened scan time without compromising the image quality for liver, spleen, and pancreas evaluations. However, these sequences should be examined before gadolinium administration when assessing the kidneys.

The Value of Second-look Ultrasound and Mammography for Assessment and Biopsy of MRI-detected Breast Lesions

Rationale and ObjectivesSuspicious lesions detected in multiparametric breast MRI can be further analyzed with second-look ultrasound (SLUS) and/or mammography. This study aims to assess the value of second-look imaging in selecting the appropriate biopsy method for different lesion characteristics. Materials and MethodsBetween January 2021 and December 2023, 212 women underwent contrast-enhanced multiparametric breast MRI at 3.0 Tesla. A total of 241 suspicious lesions (108 malignancies, 44.8%) were further assessed with SLUS and second-look mammography. Subsequent image-guided biopsy of each lesion was performed using the most suitable modality. Size-dependent lesion detection rates in SLUS and mammography were compared by means of the McNemar test. ResultsLesions referred to MRI-guided biopsy were predominantly ≤ 10 mm in size (52.8%). SLUS displayed higher detection rates than mammography for mass lesions (55.6% [95% confidence interval 46.4–64.4%] versus 16.7% [10.6–24.3%]; p<0.001) with a particularly high sensitivity for malignant mass lesions > 10 mm (88.5 % [69.9–97.6%]). In contrast, the detection rate for malignant non-mass lesions was lower in SLUS than in second-look mammography (22.0% [11.5–36.0%] versus 38.0% [24.7–52.8%]; p<0.001). The malignancy rates in ultrasound-, mammography-, and MRI-guided biopsies were 53.7%, 55.2%, and 35.0%, respectively. ConclusionSLUS is an excellent tool for further assessment and biopsy of suspicious mass lesions > 10 mm without associated calcifications. In contrast, supplemental ultrasound is of limited value in the evaluation and biopsy guidance of suspicious non-mass lesions compared to second-look mammography.

Segmentation of periapical lesions with automatic deep learning on panoramic radiographs: an artificial intelligence study

Periapical periodontitis may manifest as a radiographic lesion radiographically. Periapical lesions are amongst the most common dental pathologies that present as periapical radiolucencies on panoramic radiographs. The objective of this research is to assess the diagnostic accuracy of an artificial intelligence (AI) model based on U²-Net architecture in the detection of periapical lesions on dental panoramic radiographs and to determine whether they can be useful in aiding clinicians with diagnosis of periapical lesions and improving their clinical workflow. 400 panoramic radiographs that included at least one periapical radiolucency were selected retrospectively. 780 periapical radiolucencies in these anonymized radiographs were manually labeled by two independent examiners. These radiographs were later used to train the AI model based on U²-Net architecture trained using a deep supervision algorithm. An AI model based on the U²-Net architecture was implemented. The model achieved a dice score of 0.8 on the validation set and precision, recall, and F1-score of 0.82, 0.77, and 0.8 respectively on the test set. This study has shown that an AI model based on U²-Net architecture can accurately diagnose periapical lesions on panoramic radiographs. The research provides evidence that AI-based models have promising applications as adjunct tools for dentists in diagnosing periapical radiolucencies and procedure planning. Further studies with larger data sets would be required to improve the diagnostic accuracy of AI-based detection models.

The Evolution of Artificial Intelligence in Medical Imaging: From Computer Science to Machine and Deep Learning.

Artificial intelligence (AI), the wide spectrum of technologies aiming to give machines or computers the ability to perform human-like cognitive functions, began in the 1940s with the first abstract models of intelligent machines. Soon after, in the 1950s and 1960s, machine learning algorithms such as neural networks and decision trees ignited significant enthusiasm. More recent advancements include the refinement of learning algorithms, the development of convolutional neural networks to efficiently analyze images, and methods to synthesize new images. This renewed enthusiasm was also due to the increase in computational power with graphical processing units and the availability of large digital databases to be mined by neural networks. AI soon began to be applied in medicine, first through expert systems designed to support the clinician's decision and later with neural networks for the detection, classification, or segmentation of malignant lesions in medical images. A recent prospective clinical trial demonstrated the non-inferiority of AI alone compared with a double reading by two radiologists on screening mammography. Natural language processing, recurrent neural networks, transformers, and generative models have both improved the capabilities of making an automated reading of medical images and moved AI to new domains, including the text analysis of electronic health records, image self-labeling, and self-reporting. The availability of open-source and free libraries, as well as powerful computing resources, has greatly facilitated the adoption of deep learning by researchers and clinicians. Key concerns surrounding AI in healthcare include the need for clinical trials to demonstrate efficacy, the perception of AI tools as 'black boxes' that require greater interpretability and explainability, and ethical issues related to ensuring fairness and trustworthiness in AI systems. Thanks to its versatility and impressive results, AI is one of the most promising resources for frontier research and applications in medicine, in particular for oncological applications.

Diagnostic performance of Photon-counting CT angiography in peripheral artery disease compared to DSA as gold standard

BackgroundPhoton-counting (PC) CT has the potential to improve diagnostic confidence and image quality of CT angiography (CTA) in patients with peripheral artery disease (PAD). PurposeTo retrospectively evaluate the diagnostic performance of Photon-counting CT angiography for the assessment of stenotic disease in patients with PAD compared to digital subtraction angiography (DSA) as gold standard. Materials and MethodsAll patients undergoing PC CTA followed by DSA between November 2021 and November 2023 were included in this institutional review board approved HIPAA compliant retrospective analysis. The arterial vasculature of the lower extremity was divided into 10 segments from the iliac vasculature to the calf arterial vasculature. The images were evaluated independently by two experienced readers. Inter-reader agreement was determined using Cohen’s kappa coefficient (κ). Sensitivity, specificity, positive (PPV) and negative predictive values (NPV) as well as accuracy were calculated for PC CTA and PC pure lumen reconstruction compared to DSA as gold standard. Results109 patients (mean age 74.68 ± 11.10 years; 77 males, 32 females) were included in the retrospective analysis. PC pure lumen reconstructions was available for 91 patients (83 %). A total of 933 vascular segments for PC CTA and 780 vascular segments for PC pure lumen reconstruction were evaluated. Good to perfect inter-reader agreement was found for PC CTA (κ = 0.791) and for PC pure lumen reconstruction (κ = 0.829). Sensitivity, Specificity and accuracy for PC CTA were 91 %; 95 % and 93 %, respectively. Sensitivity, Specificity and accuracy for PC pure lumen reconstruction were 85 %, 89 % and 88 %, respectively. ConclusionPhoton-counting CTA demonstrates high sensitivity and specificity for the detection and diagnosis of stenotic lesions in PAD. PC non-calcium reconstruction does not further increase the accuracy compared to PC CTA.

Diagnosis and endodontic treatment of odontogenic cutaneous sinus tracts: A report of two cases

Introduction: The odontogenic cutaneous sinus tract (OCST), caused by pulp necrosis, creates connections between pathological spaces and anatomical structures. Extraoral manifestations pose diagnostic challenges leading to inappropriate treatments such as multiple surgery and antibiotic therapies. Effective management requires the treatment of dental pathology through endodontic therapy or extraction. This paper presents two cases of OCST and emphasizes the importance of accurate diagnosis and successful non-surgical endodontic treatment. Observations: Case report 1: A 23-year-old male patient referred by his dermatologist with a cutaneous sinus tract on his chin that had appeared five years ago. During this period, he reported spontaneous pain and pus discharge from the chin area. The examination revealed a periapical lesion associated with tooth number 31. Cone beam computed tomography (CBCT) revealed the existence of a localized perforation on the buccal alveolar table and a second root canal of the tooth. After six months of follow-up, the cutaneous sinus tract was significantly improved by the appropriate endodontic treatment. Case report 2: A 36-year-old female patient was referred by her dentist to examine a chin lesion that had appeared a month ago. In clinical examination and radiographic assessment, periapical lesion associated with two mandibular central incisor teeth #41 and #31 was observed. After identifying the source of infection, endodontic treatment was carried out, and a good improvement in fistula was observed after one month. Conclusions: The correct diagnosis and treatment of OCST are important for healing. CBCT aids to identify odontogenic origin of the sinus tract through the detection of apical periodontitis and bone lesion. Root canal therapy is the treatment of choice for these cases and should always be attempted first. Monitoring the patients is necessary until complete healing of the disease.

Knowledge, Attitude And Practice In Oral Cancer Among Dental Hygienists In Libya

Background: Oral cancer remains a global health problem. Lack of awareness leads to inadequate watchfulness regarding early signs/symptoms despite the ease of visual oral inspection. Dental hygienists play an important role in early detection of oral cancerous lesions. Evidence on the awareness and knowledge level of oral cancer and its associated risk factors among dental hygienists in Libya is scarce. The aim of this study was to assess knowledge, attitudes, and practice (KAP) regarding OC among dental hygienists in Libya. Materials and methods: Administered questionnaires were distributed on dental hygienists (through Google Forms). The questionnaire included four sections: sociodemographic information, oral cancer knowledge, attitudes toward oral cancer, and clinical practices related to oral cancer. Participants' responses to the questionnaire were analyzed using descriptive statistics. Results: Responses were received from 100 participants (60% women and 40% men). Overall, there is a good knowledge of the real risk factors among the participants (86%, 83%, and 71%) regarding the following risk factors smoking, alcohol, and poor diet respectively. The analyze related to the knowledge of non- risk factors, Just over half of the respondents correctly indicated that family history, poor oral hygiene, and poor fitting prosthesis are not risk factors for OC onset (57%; 55%; and 51%, respectively). The majority of respondents (80% &amp; 77% respectively) correctly identified leukoplakia and erythroplakia as oral potentially malignant disorders. The totality 81% of DHs believed they needed continuous updates in the future. With regard to practice, high percentage of DHs stated to perform an extra/intraoral examination and carried out palpation of the lymph nodes. Conclusions: Overall, this study highlighted the importance of introducing continuous education and training courses to dental hygienists which could increase oral cancer prevention and survival rates among oral cancer patients

Ultrasonographic liver nodules are more often benign lesions in dogs with hemoperitoneum secondary to splenic tumor rupture.

To evaluate the reliability of preoperative abdominal ultrasonography as a staging tool for dogs with hemoperitoneum due to presumed splenic tumor rupture, focusing on the detection of metastatic lesions in the liver. 99 dogs from 20 emergency and specialty hospitals across the US. Dogs with nontraumatic hemoperitoneum secondary to splenic tumor rupture were included. A post hoc analysis was conducted on data from a nationwide prospective trial investigating novel treatments for canine hemangiosarcoma. The accuracy of preoperative staging was assessed by comparing ultrasonographic findings with intraoperative observations and histologic findings. On preoperative ultrasonography, there was a 20% incidence of liver lesions identified, with no association to liver lesions seen during operation. Notably, 22% of liver lesions observed during operation were missed on preoperative ultrasonography. The presence of liver lesions on preoperative ultrasonography was associated with a higher likelihood of a benign splenic tumor diagnosis. There was no association between the identification of liver lesions on preoperative ultrasonography and the presence of metastatic disease on liver biopsy, with a sensitivity and specificity of 19% and 82%, respectively. Additionally, ultrasound had low sensitivity in detecting intra-abdominal lesions beyond the liver and spleen, with 82% of these lesions missed preoperatively. This study challenges conventional perceptions around the approach to staging in dogs with hemoperitoneum. These findings advocate for a reevaluation of the staging approach, with more comprehensive modalities like whole-body CT or MRI potentially being more warranted.

External Validation of a Previously Developed Deep Learning-based Prostate Lesion Detection Algorithm on Paired External and In-House Biparametric MRI Scans.

Purpose To evaluate the performance of an artificial intelligence (AI) model in detecting overall and clinically significant prostate cancer (csPCa)-positive lesions on paired external and in-house biparametric MRI (bpMRI) scans and assess performance differences between each dataset. Materials and Methods This single-center retrospective study included patients who underwent prostate MRI at an external institution and were rescanned at the authors' institution between May 2015 and May 2022. A genitourinary radiologist performed prospective readouts on in-house MRI scans following the Prostate Imaging Reporting and Data System (PI-RADS) version 2.0 or 2.1 and retrospective image quality assessments for all scans. A subgroup of patients underwent an MRI/US fusion-guided biopsy. A bpMRI-based lesion detection AI model previously developed using a completely separate dataset was tested on both MRI datasets. Detection rates were compared between external and in-house datasets with use of the paired comparison permutation tests. Factors associated with AI detection performance were assessed using multivariable generalized mixed-effects models, incorporating features selected through forward stepwise regression based on the Akaike information criterion. Results The study included 201 male patients (median age, 66 years [IQR, 62-70 years]; prostate-specific antigen density, 0.14 ng/mL2 [IQR, 0.10-0.22 ng/mL2]) with a median interval between external and in-house MRI scans of 182 days (IQR, 97-383 days). For intraprostatic lesions, AI detected 39.7% (149 of 375) on external and 56.0% (210 of 375) on in-house MRI scans (P < .001). For csPCa-positive lesions, AI detected 61% (54 of 89) on external and 79% (70 of 89) on in-house MRI scans (P < .001). On external MRI scans, better overall lesion detection was associated with a higher PI-RADS score (odds ratio [OR] = 1.57; P = .005), larger lesion diameter (OR = 3.96; P < .001), better diffusion-weighted MRI quality (OR = 1.53; P = .02), and fewer lesions at MRI (OR = 0.78; P = .045). Better csPCa detection was associated with a shorter MRI interval between external and in-house scans (OR = 0.58; P = .03) and larger lesion size (OR = 10.19; P < .001). Conclusion The AI model exhibited modest performance in identifying both overall and csPCa-positive lesions on external bpMRI scans. Keywords: MR Imaging, Urinary, Prostate Supplemental material is available for this article. © RSNA, 2024.

Comprehensive Review of the Utility of Dynamic Contrast-Enhanced MRI for the Diagnosis and Treatment Assessment of Spinal Benign and Malignant Osseous Disease.

Conventional MRI is currently the preferred imaging technique for detection and evaluation of malignant spinal lesions. However, this technique is limited in its ability to assess tumor viability. Unlike conventional MRI, dynamic contrast-enhanced (DCE) MRI provides insight into the physiologic and hemodynamic characteristics of malignant spinal tumors and has been utilized in different types of spinal diseases. DCE has been shown to be especially useful in the cancer setting; specifically, DCE can discriminate between malignant and benign vertebral compression fractures as well as between atypical hemangiomas and metastases. DCE has also been shown to differentiate between different types of metastases. Furthermore, DCE can be useful in the assessment of radiation therapy for spinal metastases, including the prediction of tumor recurrence. This review considers data analysis methods utilized in prior studies of DCE-MRI data acquisition and clinical implications.

MFFE CT-like MRI Sequences for the Assessment of Vertebral Fractures.

The aim of this study was to evaluate the diagnostic performance, image quality, and inter- and intra-observer agreement of the 3D T1 multi-echo fast field echo (mFFE) sequence in cervico-thoraco-lumbar vertebral fractures compared with conventional computed tomography (CT) as the gold standard. We conducted a prospective single-centre study including 29 patients who underwent spinal magnetic resonance imaging (MRI) at the surgeon's request, in addition to CT for vertebral fracture assessment and classification. A 3D T1 mFFE sequence was added to the standard MRI protocol. Consecutively, two readers analyzed the 3D mFFE sequence alone, the 3D mFFE sequence with the entire MRI protocol, including the STIR and T1 sequences, and, finally, the CT images in random order and 1 month apart. A standardized assessment was performed to determine the presence or absence of a fracture, its location, its classification according to the Genant and AO classifications for traumatic and osteoporotic fractures, respectively, the loss of height of the anterior and posterior walls of the vertebral body, and the presence of concomitant disco-ligamentous lesions. Contingency tables, intraclass correlation coefficients, and Cohen's kappa tests were used for statistical analysis. A total of 25 fractures were recorded (48% cervical, 20% thoracic, and 32% lumbar), of which 52% were classified A, according to the AO classification system. The quality of the 3D mFFE image was good or excellent in 72% of cases. Inter-observer agreement was near perfect (0.81-1) for vertebral body height and for AO and Genant classifications for all modalities. Intra-observer agreement was strong-to-near perfect between CT and the 3D mFFE sequence. Regarding the diagnostic performance of the 3D mFFE sequence, the sensitivity was 0.9200 and 0.9600, the specificity was 0.9843 and 0.9895, and the accuracy was 0.9861 and 0.9769 for Readers 1 and 2, respectively. In addition, up to 40% of intervertebral disc lesions and 33% of ligamentous lesions were detected by the 3D mFFE sequence compared to CT, allowing four AO type A fractures to be reclassified as type B. The 3D mFFE sequence allows accurate diagnosis of vertebral fractures, with superiority over CT in detecting disco-ligamentous lesions and a more precise classification of fractures, which can prompt clinicians to adapt their management despite an image quality that still requires improvement in some cases. Vertebral fractures and disco-ligamentous lesions can be assessed using CT-like MRI sequences, with 3D T1 mFFE being superior to CT for the detection of disco-ligamentous lesions. CT-like images using the 3D T1 mFFE sequence improve the diagnostic accuracy of bone structures in MRI.

Multiple myeloma: What is the most cost-effective imaging strategy for initial detection of bone lesions?

To determine the cost-effectiveness of different imaging modalities for initial detection of multiple myeloma (MM)-defining bone lesions. A Markov model from the health care system perspective for patients with MGUS was used to evaluate the incremental cost-effectiveness of five imaging techniques: skeletal survey (SS), low-dose computed tomography (LDCT), positron emission computed tomography (PETCT), and whole-body magnetic resonance imaging (WBMRI) with and without diffusion (DIFF). Model inputs, including probabilities, utilities, and costs were obtained from comprehensive literature review. Costs were estimated in 2024 U.S. dollars, effectiveness was measured in quality adjusted life years (QALYs), willingness-to-pay (WTP) threshold was set to $100,000/QALY, and timeframe of the simulation was 20years. Model analyses included Monte Carlo microsimulation and probabilistic sensitivity analysis (PSA). The most cost-effective imaging strategy was dependent on the number of patient risk factors for progression from MGUS to myeloma. At a WTP threshold of $100,000, for patients with no risk factors for progression, LDCT amassed the greatest net monetary benefit (NMB) ($1,030,913.57) while incurring the second lowest costs ($44,870.73). For patients with 1 or 2 risk factors for progression, WBMRI + DIFF amassed the greatest NMB (1 risk factor: $802,637.30, 2 risk factors: $664,430.36). WBMRI and PETCT were absolutely dominated in all cases. PSA also found that the most cost-effective strategy was dependent on the WTP threshold. Our model suggests that LDCT and WBMRI + DIFF can be the most cost-effective imaging strategies for the initial diagnosis of MM in patients, depending on the number of risk factors for progression.

The use of artificial intelligence in colonoscopic evaluations.

This review aims to highlight the transformative impact of artificial intelligence in the field of gastrointestinal endoscopy, particularly in the detection and characterization of colorectal polyps. Over the past decade, artificial intelligence has significantly advanced the medical industry, including gastrointestinal endoscopy. Computer aided diagnosis - detection (CADe) systems have shown notable success in increasing ADR. Recent meta-analyses of RCTs have demonstrated that patients undergoing colonoscopy with CADe assistance had a higher ADR compared with conventional methods. Similarly, computer aided diagnosis - characterization (CADx) systems have proven effective in distinguishing between adenomatous and nonadenomatous polyps, enhancing diagnostic confidence and supporting cost-saving measures like the resect-and-discard strategy. Despite the high performance of these systems, the variability in real-world adoption highlights the importance of integrating artificial intelligence as an assistive tool rather than a replacement for human expertise. Artificial intelligence integration in colonoscopy, through CADe and CADx systems, marks a significant advancement in gastroenterology. These systems enhance lesion detection and characterization, leading to improved diagnostic accuracy, training outcomes, and clinical workflow efficiency. While artificial intelligence offers substantial benefits, the optimal approach involves using artificial intelligence to augment the expertise of endoscopists, ensuring that clinical decisions remain under human oversight.

Low-Field Magnetic Resonance Imaging for the Detection of Medial Meniscal Lesions in Cranial Cruciate-Deficient Stifle Joints in Dogs.

The aim of this study was to substantiate the ability of noninvasive low-field magnetic resonance imaging (lfMRI) to accurately depict meniscal lesions, thereby establishing lfMRI as a viable option for the diagnosis of meniscal pathology. This was a prospective, controlled study. In each of 57 stifle joints, 5 predefined lfMRI sequences using a 0.25 T lfMRI were used. The scans were evaluated by an experienced surgeon (DECVS) and through external assessment by a radiologist (DECVDI). The patients were divided into two groups depending on their meniscal status after lfMRI. Dogs (Group 1; n = 33) without lfMRI evidence of medial meniscal pathology underwent TPLO, whereas concurrent craniomedial mini-arthrotomy was performed in patients with suspected meniscal lesions (Group 2; n = 24). Gait assessment was performed 6 weeks later and 6 months postoperatively. Cohen's kappa value of 0.8571 corresponds to a high level of agreement between the observers. A significant correlation was found between preoperative lameness severity and intraoperative meniscal damage (p < 0.04). The results showed a highly significant correlation between meniscal injuries on MRI and intraoperative findings (p < 0.0001). LfMRI appears to be a noninvasive, reliable imaging tool for the detection of medial meniscal lesions in cranial cruciate-deficient stifle joints, aiming to avoid (minimally) invasive diagnostics in healthy menisci.

Breast Lesion Detection for Ultrasound Images Using MaskFormer.

This study evaluates the performance of the MaskFormer model for segmenting and classifying breast lesions using ultrasound images, addressing ultrasound's limitations. Ultrasound used for breast cancer detection faces challenges like low image contrast and difficulty in the detection of small or multiple lesions, further complicated by variability based on operator skill. Initial experiments with U-Net and other CNN-based models revealed constraints, such as early plateauing in model loss, which indicated suboptimal learning and performance. In contrast, MaskFormer demonstrated continuous improvement, achieving higher precision in breast lesion segmentation and significantly reducing both false positives and false negatives. Comparative analysis showed MaskFormer's superior performance, with the highest precision and recall rates for malignant lesions and an overall mean average precision (mAP) of 0.943. The model's ability to detect a diverse range of breast lesions, including those potentially missed by the human eye, especially by less experienced practitioners, underscores its potential. These findings suggest that integrating AI models like MaskFormer could greatly enhance ultrasound performance for breast cancer detection, providing reliable, operator-independent image analysis and potentially improving patient outcomes on a global scale.

Revolutionizing endometriosis treatment: automated surgical operation through artificial intelligence and robotic vision.

Clinical limitations due to poverty significantly impact the lives and health of many individuals globally. Nevertheless, this challenge can be addressed with modern technologies, particularly through robotics and artificial intelligence. This study aims to address these challenges using advanced technologies in robotic surgery and artificial intelligence, proposing a method to fully automate endometriosis robotic surgery with a focus on interpretability, accuracy, and reliability. A methodology for fully automatic endometriosis surgery is introduced. Given the complexity of endometriosis lesions detection, they are categorized by their anatomical location to improve system interpretability. Then, three ensemble U-Net frameworks are designed to detect and localize common types of endometriosis lesions intraoperatively. A cross-training approach is employed, exploring U-Net models with diverse neural architectures-such as ResNet50, ResNet101, VGG19, InceptionV3, MobileNet, and EfficientNetB7-to develop U-Net ensemble models for precise endometriosis lesions segmentation. A novel image augmentation technique is also introduced, enhancing the segmentation models' accuracy and reliability. Furthermore, two U-Net models are developed to localize the ovaries and uterus, mitigating unexpected noise and bolstering the method's accuracy and reliability. The image segmentation models, assessed using the Intersection over Union (IoU) metric, achieved outstanding results: 97.57% for ovarian, 96.35% for uterine, and 92.58% for peritoneal endometriosis. This study proposes a fully automatic method for some common types of endometriosis surgery, including ovarian endometriomas and superficial endometriosis. This method is centered around three ensemble U-Net frameworks and a noise reduction technique using two additional U-Nets for localizing the ovaries and uterus. This approach has the potential to significantly improve the accuracy and reliability of robotic surgeries, potentially reducing healthcare costs and improving outcomes for patients worldwide.

Periapical lesion detection in periapical radiographs using the latest convolutional neural network ConvNeXt and its integrated models

To overcome the limitation of a single classification model’s inability to simultaneously identify multiple lesion targets within periapical radiographs, This study proposes YoCNET (Yolov5 + ConvNeXt), a novel deep learning integrated model. YoCNET leverages the target detection capability of Yolov5 and the image classification capability of ConvNeXt to achieve automatic segmentation of individual teeth and concurrent detection of periapical lesions across multiple teeth. A dataset of 1,305 periapical radiographs was used to train and validate the ConvNeXt and ResNet34 models, with an 8:2 split for training and validation. Deciduous teeth were excluded from the dataset. Furthermore, 717 individual teeth images were extracted from 200 previously unused periapical radiographs for integrated model validation. Evaluation metrics included accuracy, precision, sensitivity, F1 score, AUC (Area Under Curve), and a confusion matrix.The YoCNET integrated model demonstrated values of 90.93%, 98.88%, 85.30%, 0.9159, and 0.9757 for accuracy, precision, sensitivity, F1 score, and AUC, respectively. These metrics were superior to those achieved by the YoRNET (Yolov5 + ResNet34) integrated model, which recorded 80.47%, 83.78%, 82.16%, 0.8296, and 0.8822. The integrated model achieved high accuracy and efficiency in automatic teeh segmentation by Yolov5 and in automatically detecting multiple periapical lesions by ConvNeXt. YoCNET exhibited superior overall data performance, making it a more suitable deep learning integrated model for clinical applications.