Reproducing RECIST lesion selection via machine learning: Insights into intra and inter-radiologist variation.

Background: The Response Evaluation Criteria in Solid Tumors (RECIST) aims to provide a standardized approach to assess treatment response in solid tumors. However, discrepancies in the selection of measurable and target lesions among radiologists using these criteria pose a significant limitation to their reproducibility and accuracy. This study aimed to understand the factors contributing to this variability. Methods: Machine learning models were used to replicate, in parallel, the selection process of measurable and target lesions by two radiologists in a cohort of 40 patients from an internal pan-cancer dataset. The models were trained on lesion characteristics such as size, shape, texture, rank, and proximity to other lesions. Ablation experiments were conducted to evaluate the impact of lesion diameter, volume, and rank on the selection process. Results: The models successfully reproduced the selection of measurable lesions, relying primarily on size-related features. Similarly, the models reproduced target lesion selection, relying mostly on lesion rank. Beyond these features, the importance placed by different radiologists on different visual characteristics can vary, specifically when choosing target lesions. Worth noting that substantial variability was still observed between radiologists in both measurable and target lesion selection. Conclusions: Despite the successful replication of lesion selection, our results still revealed significant inter-radiologist disagreement. This underscores the necessity for more precise guidelines to standardize lesion selection processes and minimize reliance on individual interpretation and experience as a means to bridge existing ambiguities.

Lax Uterosacral Ligament and Urge Urinary Incontinence: MRI Findings in Symptomatic Patients Versus Healthy Volunteers.

INTRODUCTION AND HYPOTHESIS: The objective was to explore the association between urge urinary incontinence (UUI) and lax uterosacral ligaments (USL) using MRI. METHODS: Sixty-seven female participants were recruited prospectively: 41 continent volunteers (control group) and 26 patients with UUI. Static proton density- and T2-weighted turbo spin echo sequences of MR images were used. A radiologist employed a standardized grid system to record structural observations of the USLs on sequentially numbered axial MR images and then applied a four-point grading scale to assess ligament visibility. MR images were interpreted by a radiologist and a urologist, and then validated by an expert radiologist. RESULTS: The comparison between the mean length of uterosacral ligaments in the control and UUI groups was highly statistically significant (p < 0.001). The mean length of the right USL was 38 ± 11 mm, and the left USL was 35 ± 12 mm in the UUI group. In the control group, the mean length of the USL was 22 ± 9 mm on the right side and 18 ± 9 mm on the left side, along their craniocaudal extent. The highest inter-observer agreement was on the level of origin and insertion (image numbers), whereas the lowest agreement was on the anatomical site of origin and insertion of the USL in both the control and UUI groups. CONCLUSIONS: The average length of USLs in patients with UUI is significantly longer than that in healthy continent women, indicating laxity. Our findings support the relationship between the laxity of the USL and UUI symptoms and have therapeutic implications.

Magnetic resonance imaging of pilonidal sinus disease: interobserver agreement and practical MRI reporting tips.

OBJECTIVE: To evaluate the interobserver agreement for the features of natal cleft pilonidal sinus disease (PSD) on magnetic resonance imaging (MRI) and propose a standardized checklist for reporting PSD on MRI. MATERIALS AND METHODS: Forty MRI studies of 39 discrete patients with PSD were retrospectively evaluated by five independent radiologists using a standardized checklist. Fleiss' Kappa (k) coefficients of agreement were used to test the agreement between categorical variables. The MRI features of the natal cleft sepsis associated with PSD were classified into four main categories: morphology, branching and extensions, external skin openings, and the relationship of the PSD to the coccyx. A survey was created and disseminated online among general surgeons who treat patients with PSD to assess the relevance of the MRI features proposed in the standardized checklist. RESULTS: The overall agreement regarding the identification of morphology of the natal cleft sepsis was moderate (k = 0.59). Lateral and caudal extensions interobserver agreement was substantial (k = 0.64 and 0.71, respectively). However, the overall agreement regarding the individual parts of anal sphincter involved was moderate (k = 0.47). Substantial interobserver agreement was found in assessing the proximity of the PSD to the coccyx (k = 0.62). CONCLUSION: Preoperative MRI can delineate the extensions and branching of PSD with substantial agreement. MRI is superior in describing the deep extensions of PSD with better reliability than assessing the number and locations of the external openings. Expert consensus agreement is needed to establish the MRI features necessary for optimal reporting of PSD. CLINICAL RELEVANCE STATEMENT: MRI can offer valuable information about the extent of sepsis associated with pilonidal sinus disease, particularly in cases with involvement of critical anatomical structures such as the coccyx and anal triangle. MRI can potentially contribute to more accurate patient stratification and surgical planning. KEY POINTS: • The interobserver agreement for assessing PSD's lateral and caudal extension on MRI is substantial. • MRI can describe deep extensions and branching of PSD with superior reliability than assessing the number and site of external openings. • Reporting the relationship between natal cleft sepsis in PSD and the anal region may influence the surgical approach and postoperative healing.

Solitary rectal ulcer syndrome (SRUS): observational case series findings on MR defecography.

OBJECTIVE: Radiological findings in solitary rectal ulcer syndrome (SRUS) are well described for evacuation proctography (EP) but sparse for magnetic resonance defecography (MRD). In order to rectify this, we describe the spectrum of MRD findings in patients with histologically proven SRUS. MATERIALS AND METHODS: MRD from twenty-eight patients (18 female; 10 males) with histologically confirmed SRUS were identified. MRD employed a 1.5-T magnet and a standardized technique with the rectal lumen filled with gel and imaged sagittally in the supine position, before, during, and after attempted rectal evacuation. A single radiologist observer with 5 years' experience in pelvic floor imaging made the anatomical and functional measurements. RESULTS: Sixteen patients (10 female) demonstrated internal rectal intussusception and 3 patients (11%) demonstrated complete external rectal prolapse. Anterior rectoceles were noted in 12 female patients (43%). Associated anterior and middle compartment weakness (evidenced by excessive descent) was observed in 18 patients (64%). Cystocele was found in 14 patients (50%) and uterine prolapse was noted in 7 patients (25%). Enterocoeles were detected in 5 patients (18%) and peritoneocoele in 5 patients (18%). None had sigmoidocoele. Sixteen patients (57%) demonstrated delayed voiding and 13 patients (46%) incomplete voiding, suggesting defecatory dyssynergia. CONCLUSION: MRD can identify and grade both rectal intussusception and dyssynergia in SRUS, and also depict associated anterior and/or middle compartment descent. Distinction between structural and functional findings has important therapeutic implications. KEY POINTS: MRD can identify and grade both rectal intussusception and dyssynergia in patients with SRUS. MRD is an acceptable substitute to evacuation proctography in assessing anorectal dysfunctions when attempting to avoid ionizing radiation. SRUS influences the pelvic floor globally. MRD depicts associated anterior and/or middle compartment prolapse.