Mini-percutaneous nephrolithotomy vs flexible ureteroscopy for 1-2 cm lower pole renal stones: a randomised controlled trial.

OBJECTIVE: To compare the safety and efficacy of flexible ureteroscopy (f-URS) and ambulatory tubeless mini-percutaneous nephrolithotomy (mini-PCNL) in the treatment of 1-2 cm lower calyceal renal stones. PATIENTS AND METHODS: Patients who underwent f-URS and mini-PCNL for the treatment of 1-2 cm lower calyceal renal stones between October 2020 and November 2023 were evaluated in a randomised controlled trial. A total of 72 patients were included in the study. All patients underwent a computed tomography renal colic scan preoperatively, on postoperative Day 1 (POD 1), and at 3 months follow-up. We compared perioperative outcomes, including operative time and hospital stay. Additionally, we evaluated follow-up outcomes, such as the stone-free rate (SFR) and complications. All patients were discharged home on the same operative day. RESULTS: There were no significant differences in preoperative baseline data between the two surgical groups. A significantly longer median operative time was reported in the mini-PCNL group (P = 0.04). The median hospital stay was 5 h and 4 h in the mini-PCNL and f-URS groups, respectively (P = 0.14). On POD 1, the SFR, defined as the absence of residual fragments measuring 0 cm, was 50% for mini-PCNL vs 11.1% for f-URS (P < 0.001). When a total cut-off of <4 cm was utilised, the SFR was 75% in the mini-PCNL group vs 22.2% in the f-URS cohort (P < 0.001). At 3 months follow-up, the SFR remained favourable for mini-PCNL at 72.2% vs 37.1% for f-URS (P = 0.003), with a cut-off of 0 cm, and it increased to 86.1% for mini-PCNL vs 65.7% for f-URS (P = 0.04) when a total cut-off of <4 cm was applied. There was no significant difference in postoperative complications between the two groups. Two patients (5.7%) in the f-URS group required re-treatment. CONCLUSIONS: Ambulatory tubeless mini-PCNL and f-URS are effective treatment options for 1-2 cm lower calyceal renal stones. Both techniques have a comparable hospital stay and complication rates, with a significantly better SFR with mini-PCNL.

Emergency holmium laser enucleation of the prostate (HoLEP): a novel approach in the management of refractory hematuria for patients with benign prostatic hyperplasia (BPH): a single-institution experience.

INTRODUCTION: Refractory hematuria secondary to prostatic disease typically resolves with conservative management; however, this condition may require hospitalization with extensive measures to control life-threatening bleeding. The aim of this study was to report our experience using holmium laser enucleation of the prostate (HoLEP) as an emergency treatment in this clinical setting. METHODS: We conducted a retrospective review of all patients that presented to the emergency department with refractory hematuria of prostatic origin from October 2017 to September 2021, for whom hospitalization and conservative management failed to control bleeding. All emergency HoLEP procedures were performed by a single surgeon. Preoperative and intraoperative parameters, as well as perioperative outcomes, were collected and analyzed. Postoperative outcomes included duration of foley catheterization, length of postoperative hospital stay, and hospital readmissions. RESULTS: A total of 40 emergency HoLEP procedures were performed. Our cohort had a median prostate volume of 110.5 cc and a median resected weight of 81 g. Twenty-seven patients (67.5%) were on anticoagulant or antiplatelet medications on admission. The urethral catheter was removed within 1 day in 95% of patients with a successful trial of void (TOV). Moreover, 92.5% of patients were discharged home within 24 h of their procedure. Two patients (5%) experienced clot retention within one-week post-discharge with a 2.5% overall readmission rate. All postoperative parameters, including International Prostate Symptom Score (IPSS), quality of life (QoL), maximum flow rate (Qmax), and post-void residual volume (PVR), showed significant improvement at 1 year follow up. CONCLUSION: Our experience demonstrates that emergency HoLEP is an effective treatment option for patients with refractory hematuria of prostatic origin. Further studies are warranted to consolidate our results.

Radiologist Productivity Analytics: Factors Impacting Abdominal Pelvic CT Exam Reporting Times.

The purpose is to determine factors impacting radiologist abdominal pelvic CT exam reporting time. This study was Research Ethics Board approved. Between January 2019 and March 2020, consecutive abdominal pelvic CT exams were documented as structured or unstructured based on application of templates with separate sections for different organs or organ systems. Radiologist reporting location, patient class (inpatient, Emergency Department (ED) patient, outpatient), radiologist fellowship-training, report word count, and radiologist years of experience were documented. Median reporting times were compared using the Wilcoxon Rank-sum test, Kruskal-Wallis test, and regression analysis. Spearman's rank correlation was used to determine correlation between word count and radiologist experience with reporting time. P < 0.05 is defined statistical significance. A total of 3602 abdominal pelvic CT exam reports completed by 33 radiologists were reviewed, including 1150 outpatient and 2452 inpatient and Emergency Department (ED) cases. 1398 of all reports were structured. Median reporting time for structured and unstructured reports did not differ (P = 0.870). Reports dictated in-house were completed faster than reports dictated remotely (P < 0.001), and reports for inpatients/ED patients were completed faster than for outpatients (P < 0.001). Reporting time differences existed between radiologists (P < 0.001) that were not explained by fellowship training (P = 0.762). Median reporting time had a weak correlation with word count (ρ = 0.355) and almost no correlation with radiologist years of experience (ρ = 0.167), P < 0.001. Abdominal pelvic CT reporting is most efficient when dictations are completed in-house and for high-priority cases; the use of structured templates, radiologist fellowship training, and years of experience have no impact on reporting times.

Safety and efficacy of ambulatory tubeless mini-percutaneous nephrolithotomy in the management of 10-25 mm renal calculi A retrospective study.

INTRODUCTION: This study aimed to assess the safety and efficacy of ambulatory minipercutaneous nephrolithotomy (mini-PCNL) in a totally tubeless exit (without a nephrostomy tube or an internal stent) and tubeless exit (without a nephrostomy tube but with an internal stent) for the treatment of renal calculi 10-25 mm in size. METHODS: We conducted a retrospective analysis of patients who underwent mini-PCNL at our institution between September 2018 and September 2022. The study included a cohort of 95 patients diagnosed with renal calculi measuring 10-25 mm. All patients underwent a computed tomography (CT) renal colic scan preoperatively, on postoperative day one (POD 1), and at three-month followup. Patient demographics and outcome parameters were recorded, including stone characteristics, operative time, hospital stay, stone-free rate (SFR), complication rates, and subsequent emergency room (ER) visits. Patients were considered stone-free if they had no fragments or residual fragments measuring <4 mm. RESULTS: The median maximum stone diameter was 16 mm (10-25 mm). Twenty-nine patients (30.5%) had multiple renal calculi. The median operative time was 64 (38-135) minutes. Eighty-six patients (90.5%) underwent a totally tubeless procedure, without a nephrostomy tube or an internal stent. All patients were discharged home on the same operative day with a median hospitalization time of six hours. Seven (7.4%) postoperative ER visits were recorded, and two (2.1%) led to hospital readmission. The frequency of grade I, II, and III Clavien-Dindo complications were 18 (18.9%), one (1.1%), and one (1.1%), respectively. The SFR on POD 1 and three-month followup was 73.7% and 92.6%, respectively. Two patients in the study required retreatment. CONCLUSIONS: Ambulatory tubeless mini-PCNL is a safe and effective treatment option for 10-25 mm renal stones. Experienced institutions can safely adopt ambulatory mini-PCNL as a treatment option without an increased risk of postoperative complications, ER visits, or hospital readmissions.

Factors predicting stone-free rates after retrograde intrarenal surgery for lower pole kidney stones: A single-center, retrospective analysis.

INTRODUCTION: We aimed to investigate the factors impacting stone clearance following retrograde intrarenal surgery (RIRS) for lower pole kidney stones and to determine whether there is a significant relationship between the infundibular pelvic angle (IPA) of the kidney's lower pole and stone fragment clearance. METHODS: We retrospectively reviewed patients who underwent flexible ureteroscopy (f-URS) for lower pole renal calculi between December 2020 and July 2023 at our institution. Patient demographics and stone parameters were recorded, including stone size, number, volume, density, and IPA. Intraoperative data, including total operative time, lasing time, type of laser used, and stone composition, were collected and analyzed. All patients underwent a computed tomography (CT) scan at three months followup. We recorded the presence of residual stones and the percentage of stone volume reduction. Patients with a stone size ≤3 mm were deemed stone-free. All patients were discharged home on the same operative day. RESULTS: A total of 123 patients were included in the study: 71 in the stone-free group (group 1) and 52 in the residual stones group (group 2). On univariate analysis, there were significant differences between the two groups in terms of stone size, IPA, and the type of ureteroscopy used. At three months followup, 96% (24/25) of patients with an IPA <30° had residual stones, compared to 28.6% (28/98) of patients with an IPA >30° (p<0.001). There was no significant difference in the intraoperative or postoperative complications between the two groups. On multivariate analysis, IPA and stone size were the only predictive factors for the presence of residual stones. Twelve patients (23.1%) from group 2 required retreatment. CONCLUSIONS: RIRS is an effective treatment option for the management of lower pole kidney stones. IPA, in conjunction with stone size, appears to dictate the stone clearance rates of RIRS for lower pole stones.

Use of downstream stress imaging tests for risk stratification of patients presenting to the emergency department with chest pain and low HEART score.

BACKGROUND: Patients with low HEART (History, Electrocardiogram, Age, Risk factors, and Troponin level) risk scores who are discharged from the emergency department (ED) may present clinical challenges and diagnostic dilemmas. The use of downstream non-invasive stress imaging (NISI) tests in this population remains uncertain. Therefore, this study aims to investigate the value of NISI in risk stratification and predicting cardiac events in patients with low-risk HEART scores (LRHSs). METHODS: We prospectively included 1384 patients with LRHSs between March 2019 and March 2021. All the patients underwent NISI (involving myocardial perfusion imaging/stress echocardiography). The primary endpoints included cardiac death, non-fatal myocardial infarction and unplanned coronary revascularisation. Secondary endpoints encompassed cardiovascular-related admissions or ED visits. RESULTS: The mean patient age was 64±14 years, with 670 (48.4%) being women. During the 634±104 days of follow-up, 58 (4.2%) patients experienced 62 types of primary endpoints, while 60 (4.3%) developed secondary endpoints. Multivariable Cox models, adjusted for clinical and imaging variables, showed that diabetes (HR: 2.38; p=0.008), HEART score of 3 (HR: 1.32; p=0.01), history of coronary artery disease (HR: 2.75; p=0.003), ECG changes (HR: 5.11; p<0.0001) and abnormal NISI (HR: 16.4; p<0.0001) were primary endpoint predictors, while abnormal NISI was a predictor of secondary endpoints (HR: 3.05; p<0.0001). CONCLUSIONS: NISI significantly predicted primary cardiac events and cardiovascular-related readmissions/ED visits in patients with LRHSs.

Impact of pre-operative abdominal MRI on survival for patients with resected pancreatic carcinoma: a population-based study.

Background: This study determined the impact of pre-operative abdominal MRI on all-cause mortality for patients with resected PDAC. Methods: All adult (≥18 years) PDAC patients who underwent pancreatectomy between January 2011 and December 2022 in Ontario, Canada, were identified for this population-based cohort study (ICD-O-3 codes: C250, C251, C252, C253, C257, C258). Patient demographics, comorbidities, PDAC stage, medical and surgical management, and survival data were sourced from multiple linked provincial administrative databases at ICES. All-cause mortality was compared between patients with and without a pre-operative abdominal MRI after controlling for multiple covariates. Findings: A cohort of 4579 patients consisted of 2432 men (53.1%) and 2147 women (46.9%) with a mean age of 65.2 years (standard deviation: 11.2 years); 2998 (65.5%) died while 1581 (34.5%) survived. Median follow-up duration post-resection was 22.4 months (interquartile range: 10.8-48.8 months), and median survival post-pancreatectomy was 25.9 months (95% confidence interval [95% CI]: 24.8, 27.5). Patients who underwent a pre-operative abdominal MRI had a median survival of 33.1 months (95% CI: 30.7, 37.2) compared to 21.1 months (95% CI: 19.8, 22.6) for all others. A total of 2354/4579 (51.4%) patients underwent a pre-operative abdominal MRI, which was associated with a 17.2% (95% CI: 11.0, 23.1) decrease in the rate of all-cause mortality, with an adjusted hazard ratio (aHR) of 0.828 (95% CI: 0.769, 0.890). Interpretation: Pre-operative abdominal MRI was associated with improved overall survival for PDAC patients who underwent pancreatectomy, possibly due to better detection of liver metastases than CT. Funding: Northern Ontario Academic Medicine Association (NOAMA) Clinical Innovation Fund.

Endophytic to total tumour volume ratio: An added variable to patients with T1b/T2 renal tumours undergoing partial nephrectomy.

INTRODUCTION: Partial nephrectomy is the standard of care to patients with small renal masses. It is still encouraged to larger tumours whenever feasible. The aim of this study is to look for the endophytic to total tumour volume ratio as an added variable to study the complexity of partial nephrectomy to patients with T1b/ T2 renal tumours. METHODS: Retrospective data collection of patients that had partial nephrectomy for T1b/T2 renal tumours by a single surgeon was done. Radiological re-assessment for the CT images to measure the endophytic to total tumour volume ratio was done. RESULTS: The mean age of the patients was 63 years. The study included 25 males and 11 females. All cases were managed by open surgery using retroperitoneal transverse lateral lumbotomy and warm ischemia was used in all patients. The mean tumour volume was 74 cc, the mean endophytic tumour volume was 29 cc. The mean percentage of endophytic to total tumour volume was 42%. CONCLUSIONS: Partial nephrectomy is safe for most of the patients with good performance status, having large renal masses. More complex surgery can be predicted in patients with endophytic to total tumour volume greater than 42%.

Thulium Fiber Laser vs Pulse-Modulated Holmium MOSES Laser in Flexible Ureteroscopy for the Management of Kidney Stones: A Single-Center Retrospective Analysis.

Introduction and Objective: The study's primary objective was to compare the laser efficiency and clinical outcomes of two widely used systems, the holmium MOSES laser and the thulium fiber laser (TFL), in managing kidney stones. The secondary outcomes were to evaluate the impact of stone composition on laser efficacy. Methods: We conducted a retrospective review of patients who underwent flexible ureteroscopy (f-URS) for solitary renal calculi between December 2020 and August 2022 at our institution and had a 3-month postoperative CT scan. Patient demographics and stone parameters were recorded, including stone site, size, volume, and density. Intraoperative data were collected and analyzed, including total operative time, ureteroscopy time, lasing time, technique, total energy delivered, and stone composition. All patients underwent a CT scan at 3 months follow-up. We recorded the presence of residual stones and the percentage of stone volume reduction. Ablation efficiency was calculated by dividing the energy utilized (J) by the stone volume (mm3). The ablation speed was calculated by dividing the stone volume (mm3) by the lasing time (seconds). Patients with a stone size <4 mm were deemed stone-free. Results: The MOSES and TFL groups comprised 62 and 49 patients, respectively. There were no significant differences between groups for baseline patient demographics or stone characteristics. The two modalities had comparable total energy, laser time, efficacy, and ablation speeds. No differences were detected in stone-free rates or complications between both groups. When dealing with calcium phosphate stones, we observed that the lasing time was significantly shorter with MOSES than TFL (7.95 vs 10.85 minutes, respectively [p = 0.01]). Conclusions: MOSES and TFL laser systems had comparable efficacy for lithotripsy of renal calculi during f-URS; however, calcium phosphate stones had a longer lasing time with TFL. REB Number: 100210.

Same-day trial of void and discharge following standard vs. MOSESTM holmium laser enucleation of the prostate: A single-center experience.

INTRODUCTION: We aimed to compare perioperative and postoperative outcomes and to assess the safety and feasibility of same-day trial of void (TOV) in patients who underwent standard holmium laser enucleation of the prostate (HoLEP) vs. MOSESTM HoLEP (MoLEP). METHODS: We conducted a retrospective review of prospectively collected data of patients that underwent HoLEP (100 W) or MoLEP (120 W) with same-day catheter removal three hours postoperatively at our institution from August 2018 to September 2021. Patient demographics, intraoperative parameters, and postoperative outcomes were analyzed. Data were compared as means with standard deviation and medians with interquartile range (IQR) or numbers and percentages. Continuous and categorical variables were assessed using the Mann-Whitney U test and Chi-squared test, respectively. Predictors of shorter enucleation time and failed same-day TOV were investigated. RESULTS: Of the 90 patients included, 28 underwent HoLEP while 62 had MoLEP. There was no significant difference between the groups in terms of the successful TOV (23 [82%] vs. 58 [93.5%], p=0.1) and readmission rate (3 [10.7%] vs. 1 [1.6%], p=0.08); however, the MoLEP group had a significantly shorter mean enucleation time (p<0.001), mean hemostasis time (p<0.001), mean morcellation time (p=0.003), and lower mean energy used (p<0.001). On the logistic regression model, MoLEP (odds ratio [OR] 0.03, 95% confidence interval [CI] 0.007-0.19, p<0.001), lower preoperative prostate-specific antigen (PSA) test (OR 1.25, 95% CI 1.01-1.55, p=0.03), and smaller prostate size (OR 1.06, 95% CI 1.02-1.09, p<0.001) were independent predictors of shorter enucleation time. History of preoperative retention was the only significant factor associated with a failed same-day TOV (p=0.04). There was no difference in intraoperative or postoperative complication rates or postoperative functional outcomes between the two technologies. CONCLUSIONS: Same-day TOV and discharge are feasible following standard HoLEP and MoLEP, with comparable outcomes; however, the use of MOSESTM technology offered better enucleation efficiency with excellent hemostatic potential. Preoperative retention was the only predictor of failed same-day TOV.

A case of splenic artery aneurysm and rupture in a patient on a vascular endothelial growth factor inhibitor for renal cell carcinoma.

BACKGROUND: Pazopanib is a vascular endothelial growth factor inhibitor that is used in the treatment of metastatic renal cell carcinoma. Post market reports demonstrate an increasing awareness of the association of arterial aneurysms and dissections with vascular endothelial growth factor inhibitor use, although, few reports exist for pazopanib. CASE: Here we report a 69-year-old patient with minimal cardiovascular risk factors who developed a rupture of a splenic arterial aneurysm after more than 5 years of effective treatment with pazopanib for metastatic renal cell carcinoma. CONCLUSION: This case report outlines the necessity to monitor patients while on pazopanib, even when there are minimal risk factors and long periods of tolerance.

Developing a machine learning model to predict patient need for computed tomography imaging in the emergency department.

Overcrowding is a well-known problem in hospitals and emergency departments (ED) that can negatively impact patients and staff. This study aims to present a machine learning model to detect a patient's need for a Computed Tomography (CT) exam in the emergency department at the earliest possible time. The data for this work was collected from ED at Thunder Bay Regional Health Sciences Centre over one year (05/2016-05/2017) and contained administrative triage information. The target outcome was whether or not a patient required a CT exam. Multiple combinations of text embedding methods, machine learning algorithms, and data resampling methods were experimented with to find the optimal model for this task. The final model was trained with 81, 118 visits and tested on a hold-out test set with a size of 9, 013 visits. The best model achieved a ROC AUC score of 0.86 and had a sensitivity of 87.3% and specificity of 70.9%. The most important factors that led to a CT scan order were found to be chief complaint, treatment area, and triage acuity. The proposed model was able to successfully identify patients needing a CT using administrative triage data that is available at the initial stage of a patient's arrival. By determining that a CT scan is needed early in the patient's visit, the ED can allocate resources to ensure these investigations are completed quickly and patient flow is maintained to reduce overcrowding.

TB-Net: A Tailored, Self-Attention Deep Convolutional Neural Network Design for Detection of Tuberculosis Cases From Chest X-Ray Images.

Tuberculosis (TB) remains a global health problem, and is the leading cause of death from an infectious disease. A crucial step in the treatment of tuberculosis is screening high risk populations and the early detection of the disease, with chest x-ray (CXR) imaging being the most widely-used imaging modality. As such, there has been significant recent interest in artificial intelligence-based TB screening solutions for use in resource-limited scenarios where there is a lack of trained healthcare workers with expertise in CXR interpretation. Motivated by this pressing need and the recent recommendation by the World Health Organization (WHO) for the use of computer-aided diagnosis of TB in place of a human reader, we introduce TB-Net, a self-attention deep convolutional neural network tailored for TB case screening. We used CXR data from a multi-national patient cohort to train and test our models. A machine-driven design exploration approach leveraging generative synthesis was used to build a highly customized deep neural network architecture with attention condensers. We conducted an explainability-driven performance validation process to validate TB-Net's decision-making behavior. Experiments on CXR data from a multi-national patient cohort showed that the proposed TB-Net is able to achieve accuracy/sensitivity/specificity of 99.86/100.0/99.71%. Radiologist validation was conducted on select cases by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed consistency between radiologist interpretation and critical factors leveraged by TB-Net for TB case detection for the case where radiologists identified anomalies. The proposed TB-Net not only achieves high tuberculosis case detection performance in terms of sensitivity and specificity, but also leverages clinically relevant critical factors in its decision making process. While not a production-ready solution, we hope that the open-source release of TB-Net as part of the COVID-Net initiative will support researchers, clinicians, and citizen data scientists in advancing this field in the fight against this global public health crisis.

COVID-Net CXR-2: An Enhanced Deep Convolutional Neural Network Design for Detection of COVID-19 Cases From Chest X-ray Images.

As the COVID-19 pandemic devastates globally, the use of chest X-ray (CXR) imaging as a complimentary screening strategy to RT-PCR testing continues to grow given its routine clinical use for respiratory complaint. As part of the COVID-Net open source initiative, we introduce COVID-Net CXR-2, an enhanced deep convolutional neural network design for COVID-19 detection from CXR images built using a greater quantity and diversity of patients than the original COVID-Net. We also introduce a new benchmark dataset composed of 19,203 CXR images from a multinational cohort of 16,656 patients from at least 51 countries, making it the largest, most diverse COVID-19 CXR dataset in open access form. The COVID-Net CXR-2 network achieves sensitivity and positive predictive value of 95.5 and 97.0%, respectively, and was audited in a transparent and responsible manner. Explainability-driven performance validation was used during auditing to gain deeper insights in its decision-making behavior and to ensure clinically relevant factors are leveraged for improving trust in its usage. Radiologist validation was also conducted, where select cases were reviewed and reported on by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed that the critical factors leveraged by COVID-Net CXR-2 are consistent with radiologist interpretations.

Partial Nephrectomy for T1b/T2 Renal Mass: An Added Shift from Radical Nephrectomy.

The aim of our study was to show our short-term experience in managing large renal masses (cT1b/T2) through partial nephrectomy (PN) over the last 3 years. Retrospective data collection for all patients managed by PN for renal masses larger than 4 cm over the last 3 years. Epidemiological data were collected. Surgical data including surgical and ischemic times as well as intra and postoperative complications were collected. Pre- and postoperative estimated glomerular filtration rate (eGFR) data were collected and correlated as well as postoperative complications and recurrence. We could identify 47 patients managed by PN for radiologically confirmed >4 cm renal masses. The mean age of the patients was 55.7 ± 13.4, including 29 males and 18 females. Masses were T1b and T2 in 40 and 7 patients, respectively. The mean tumor size was 6.2 ± 1.5 cm. Using renal nephrometry score; 8, 28, and 11 had low, moderate, and high complexity, respectively. Renal cell carcinoma (RCC) was identified in 42 patients. Five patients out of 42 cancerous cases (12%) had pathological T3 RCC. The mean preoperative and postoperative eGFR were 89.09 ± 12.41 and 88.50 ± 10.50, respectively (P 0.2). The median follow-up was 14 months and within that short time, no patient had evidence for cancer recurrence. PN for large renal masses is safe in experienced hands and should be attempted in a higher percentage of patients, regardless of the tumor complexity. No cancer recurrence or deterioration of renal function was observed within our short-term follow-up.

Magnetic Resonance Imaging Features in Different Types of Invasive Breast Cancer: A Systematic Review of the Literature.

Breast cancer is the most common malignancy affecting women worldwide, and early diagnosis of breast cancer is the key to its successful and effective treatment. Traditional imaging techniques such as mammography and ultrasound are used to detect and configure breast abnormalities; unfortunately, these modalities have low sensitivity and specificity, particularly in young patients with dense breast tissue, breast implants, or post-surgical scar/architecture distortions. Therefore, breast magnetic resonance imaging (MRI) has been superior in the characterization and detection of breast cancer, especially that with invasive features. This review article explores the importance of breast MRI in the early detection of invasive breast cancer versus traditional tools, including mammography and ultrasound, while also analyzing the use of MRI as a screening tool for high-risk women. We will also discuss the different MRI features for invasive ductal carcinoma and lobular carcinoma and the role of breast MRI in the detection of ductal carcinoma in situ with a focus on the utilization of new techniques, including MR spectroscopy and diffusion-weighted imaging.

MEDUSA: Multi-Scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis.

Medical image analysis continues to hold interesting challenges given the subtle characteristics of certain diseases and the significant overlap in appearance between diseases. In this study, we explore the concept of self-attention for tackling such subtleties in and between diseases. To this end, we introduce, a multi-scale encoder-decoder self-attention (MEDUSA) mechanism tailored for medical image analysis. While self-attention deep convolutional neural network architectures in existing literature center around the notion of multiple isolated lightweight attention mechanisms with limited individual capacities being incorporated at different points in the network architecture, MEDUSA takes a significant departure from this notion by possessing a single, unified self-attention mechanism with significantly higher capacity with multiple attention heads feeding into different scales in the network architecture. To the best of the authors' knowledge, this is the first "single body, multi-scale heads" realization of self-attention and enables explicit global context among selective attention at different levels of representational abstractions while still enabling differing local attention context at individual levels of abstractions. With MEDUSA, we obtain state-of-the-art performance on multiple challenging medical image analysis benchmarks including COVIDx, Radiological Society of North America (RSNA) RICORD, and RSNA Pneumonia Challenge when compared to previous work. Our MEDUSA model is publicly available.