Radiology Training Research Articles

When deep learning is not enough: artificial life as a supplementary tool for segmentation of ultrasound images of breast cancer.

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape- and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H3 = 0.26; the medium complexity level, Dice = 0.91 and H3 = 0.82; and the hardest complexity level, Dice = 0.90 and H3 = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H3 = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H3 = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H3 = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http://tinyurl.com/mr4ah687 .

How Can We Identify Extraprostatic Extension (EPE) Before Surgery? The Use of a Preoperative Prostate MRI EPE Scoring System to Assess Postprostatectomy Locally Advanced Prostate Cancer.

Background: Distinguishing between organ-confined disease and extraprostatic extension (EPE) is crucial for the treatment of patients with prostate cancer. EPE is associated with an increased risk of biochemical recurrence, positive surgical margins, and metastatic disease. An MRI-based EPE scoring system was developed by Mehralivand in 2019; however, it has not been adopted in the Urology community. The purpose of this study is to evaluate the association of MRI-based EPE scoring with the pathologic EPE (pEPE) after radical prostatectomy. Methods: We conducted a retrospective review on a prospectively collected database of male patients who underwent a prostate MRI with EPE scoring by a trained genitourinary radiologist and subsequent robotic radical prostatectomy at our institution from September 2020 to December 2022. The associations between MRI EPE (mEPE) score and the presence of EPE on surgical pathology (pEPE) were examined using multivariable logistic regression. Results: A total of 194 patients met inclusion criteria with a median age of 63 years and prostate specific antigen (PSA) 7 ng/mL. Among those with mEPE score 3, 96% had pEPE. Those patients with an mEPE score ≥2 had an increased risk of pEPE compared with those with mEPE score 0 (odds ratio 3.79; 95% confidence interval 1.28-11.3) Furthermore, those with an mEPE score 3 were significantly more likely to have pEPE compared with those with mEPE score 0, 1 and 2 independently. Conclusion: MRI EPE is a straightforward tool that strongly correlates with the presence of pEPE. If validated prospectively, this scoring system could assist in counseling patients regarding nerve-sparing approach.

Designing a Collaborative Breast Radiology Training Program to Tackle Tanzania’s Breast Cancer Crisis

Breast cancer incidence and mortality continue to increase in Africa. In Tanzania, breast cancer is the second leading cause of cancer death for women, and breast cancer incidence and mortality are projected to increase by 80% by 2030. Education gaps among health care workers, delayed presentation, limited screening, and low health literacy all pose significant challenges to providing optimal breast cancer care. Considering these factors, it is imperative to train expert breast radiologists. We present a collaborative breast radiology training program in Tanzania aimed at building local capacity to address national breast cancer challenges. Although cancer control in Africa faces many challenges, developing customized training programs for breast radiology, in conjunction with national cancer programs, may represent a key strategy for addressing breast cancer.

Radiology advocacy: Promoting collaboration between trainees and professional societies

From mammographic screening guidelines to resident work hour regulations, public policy affects every aspect of the practice of radiology and ultimately determines how radiological care is delivered to patients. Shaping public policy through advocacy is therefore critical to ensure patient access to equitable, high-quality radiological care. In advocacy, individual practicing radiologists and radiology trainees can increase the scope of their influence by collaborating with professional radiology societies. When radiology trainees participate in organized radiology advocacy, they learn about regulatory and legislative issues that will affect their careers, and they learn how to effect policy change. Radiology societies in turn benefit from trainee involvement, as engaging trainees early in their careers leads to more robust future participation and leadership. To encourage trainee involvement, radiology societies can engage individual residency programs and medical student radiology interest groups, invest in trainee-focused events, and maximize the number of positions of responsibility open to trainees. To circumvent the barriers to participation that many trainees face, radiology societies can make meeting proceedings free and available through virtual mediums. Through active collaboration, trainees and professional societies can help assure a bright future for radiologists and patients in need of radiological care.

Investigation of candidates for reactor produced radioactive materials in support of radiological training exercises

Investigation of candidates for reactor produced radioactive materials in support of radiological training exercises

Can gas and infection coexist in the intervertebral disc? A retrospective analysis of percutaneously biopsied suspected discitis-osteomyelitis cases.

To retrospectively evaluate the correlation between intradiscal gas and infection in patients percutaneously biopsied for suspected discitis-osteomyelitis. We retrospectively reviewed all CT-guided discitis-osteomyelitis biopsies performed between 2002 and 2022. Two independent trained musculoskeletal radiologists evaluated for presence of gas on CT and/or MRI within 1week of the biopsy. Disagreements were resolved by a third musculoskeletal radiologist. CT was considered the gold standard for the detection of intradiscal gas. Pathology, microbiology, and imaging and clinical follow-up were used as the gold standard for presence of infection. Interrater agreement on CT and MRI, sensitivity, and positive predictive value were calculated, using the presence of gas as an indicator (test positive) for "no infection." There were 284 biopsies in 275 subjects (mean age 58 ± 1.0 (range 4-99) years; 101 (37%) females and 174 (63%) males). Of the biopsies, 12 (4%) were cervical, 80 (28%) were thoracic, 192 (68%) were lumbar, and 200 (70%) were considered true discitis-osteomyelitis based on pathology, imaging, and clinical follow-up. Interrater agreement was excellent for CT (kappa = 0.83) and poor for MRI (kappa = - 0.021). The presence of gas had a 94% specificity and 76% negative predictive value for the absence of infection. CT is the preferred method for detecting intradiscal gas. The presence of gas means that discitis-osteomyelitis is unlikely. If intradiscal gas is present in the setting of discitis-osteomyelitis, the gas bubbles tend to be smaller and fewer in number.

Application of Artificial Intelligence (AI) System in Opportunistic Screening and Diagnostic Population in a Middle-income Nation.

This study evaluates the effectiveness of artificial intelligence (AI) in mammography in a diverse population from a middle-income nation and compares it to traditional methods. A retrospective study was conducted on 543 mammograms of 467 Malays, 48 Chinese, and 28 Indians in a middle-income nation. Three breast radiologists interpreted the examinations independently in two reading sessions (with and without AI support). Breast density and BI-RADS categories were assessed, comparing the accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) results. Of 543 mammograms, 69.2% had lesions detected. Biopsies were performed on 25%(n=136), with 66(48.5%) benign and 70(51.5%) malignant. Substantial agreement in density assessment between the radiologist and AI software (κ =0.606, p < 0.001) and the BI-RADS category with and without AI (κ =0.74, p < 0.001). The performance of the AI software was comparable to the traditional methods. The sensitivity, specificity, PPV, and NPV or radiologists alone, radiologist + AI, and AI alone were 81.9%,90.4%,56.0%, and 97.1%; 81.0%, 93.1%,55.5%, and 97.0%; and 90.0%,76.5%,36.2%, and 98.1%, respectively. AI software enhances the accuracy of lesion diagnosis and reduces unnecessary biopsies, particularly for BI-RADS 4 lesions. The AI software results for synthetic were almost similar to the original 2D mammography, with AUC of 0.925 and 0.871, respectively. AI software may assist in the accurate diagnosis of breast lesions, enhancing the efficiency of breast lesion diagnosis in a mixed population of opportunistic screening and diagnostic patients. • The use of artificial intelligence (AI) in mammography for population-based breast cancer screening has been validated in high-income nations, with reported improved diagnostic performance. Our study evaluated the usage of an AI tool in an opportunistic screening setting in a multi-ethnic and middle-income nation. • The application of AI in mammography enhances diagnostic accuracy, potentially leading to reduced unnecessary biopsies. • AI integration into the workflow did not disrupt the performance of trained breast radiologists, as there is a substantial inter-reader agreement for BI-RADS category assessment and breast density.

Augmented Reality in Interventional Radiology: Transforming Training Paradigms.

The ascent of medical technology places augmented reality (AR) at the forefront of potential advancements in interventional radiology (IR) training. This review delves into the symbiotic relationship between AR and conventional IR training, casting light on the opportunities and hurdles intrinsic to this integration. A targeted literature review was conducted using the databases PubMed, Cochrane Library, and Embase. Search terms included ((("Augmented Reality" OR "Virtual Reality")) AND ((Education OR Training))) AND (("Interventional Radiology")). Ten studies identified using the comprehensive inclusion criteria helped scrutinize the use of AR in IR training. Key outcomes include improved procedural accuracy, reduced training duration, and heightened trainee confidence. However, it also identifies limitations such as small sample sizes, reliance on simulation environments, and technological constraints in AR implementation. Despite these challenges, the review underscored the transformative potential of AR in IR education, suggesting its capacity to revolutionize training methodologies. However, it also calls for continued technological development and empirical research to address current challenges and fully leverage AR's capabilities in medical education.

Deep-learning for automated detection of MSU deposits on DECT: evaluating impact on efficiency and reader confidence.

Dual-energy CT (DECT) is a non-invasive way to determine the presence of monosodium urate (MSU) crystals in the workup of gout. Color-coding distinguishes MSU from calcium following material decomposition and post-processing. Manually identifying these foci (most commonly labeled green) is tedious, and an automated detection system could streamline the process. This study aims to evaluate the impact of a deep-learning (DL) algorithm developed for detecting green pixelations on DECT on reader time, accuracy, and confidence. We collected a sample of positive and negative DECTs, reviewed twice-once with and once without the DL tool-with a 2-week washout period. An attending musculoskeletal radiologist and a fellow separately reviewed the cases, simulating clinical workflow. Metrics such as time taken, confidence in diagnosis, and the tool's helpfulness were recorded and statistically analyzed. We included thirty DECTs from different patients. The DL tool significantly reduced the reading time for the trainee radiologist (p = 0.02), but not for the attending radiologist (p = 0.15). Diagnostic confidence remained unchanged for both (p = 0.45). However, the DL model identified tiny MSU deposits that led to a change in diagnosis in two cases for the in-training radiologist and one case for the attending radiologist. In 3/3 of these cases, the diagnosis was correct when using DL. The implementation of the developed DL model slightly reduced reading time for our less experienced reader and led to improved diagnostic accuracy. There was no statistically significant difference in diagnostic confidence when studies were interpreted without and with the DL model.

Preoperative Strategies for Locally Advanced Colon Cancer.

Neoadjuvant chemotherapy is safe for patients with locally advanced colon cancer (LACC). The FOxTROT trial demonstrated a reduction in residual and recurrent cancer at 2 years with neoadjuvant chemotherapy for patients with cT3-4 LACC. Preoperative chemotherapy should be avoided, if possible, for patients with dMMR LACC, as over 50% of dMMR cancers have no pathologic response. Early universal testing of MMR status is critical to selecting the appropriate neoadjuvant therapy. Concerns about CT staging of LACC have limited uptake of neoadjuvant chemotherapy, as approximately 25% of patients with cT3-T4 cancer on CT have low-risk stage II disease. Development of CT criteria for malignant nodes should reduce the risk of over-staging. A multidisciplinary approach is needed to identify patients for neoadjuvant therapy. Neoadjuvant immunotherapy is safe and results in dramatic pathologic responses in patients with dMMR LACC. Longer follow-up is needed to determine if the exceptionally high pathologic response rates observed will translate into long-term remission. Remarkably, neoadjuvant immunotherapy has been found to cause major pathologic responses in a subset of patients with pMMR LACC, indicating the potential to cure more patients with this common cancer. Patients with cT4 LACC, whether stage II or III, have a substantial risk of recurrence despite adjuvant fluoropyrimidine plus oxaliplatin chemotherapy. We recommend neoadjuvant systemic therapy for all patients with cT4b LACC (dMMR and pMMR). Features of T4b disease are routinely reported by radiology. We use three cycles of FOLFOX chemotherapy for patients with cT4b pMMR LACC, due to the high rate of compliance and improvement in residual and recurrent disease. Patients with cT4b dMMR LACC should receive neoadjuvant immunotherapy, if there are no contraindications. Clinical trials of neoadjuvant therapy for LACC are of great interest and should provide training for radiologists to identify eligible patients. Results are anticipated from multiple ongoing trials of neoadjuvant chemotherapy, immunotherapy, and targeted therapy for pMMR LACC and immunotherapy for dMMR LACC.

Expanding radiology training: lessons from establishing a new Dorset training programme

Expanding radiology training: lessons from establishing a new Dorset training programme

Exploring the performance and explainability of fine-tuned BERT models for neuroradiology protocol assignment

BackgroundDeep learning has demonstrated significant advancements across various domains. However, its implementation in specialized areas, such as medical settings, remains approached with caution. In these high-stake environments, understanding the model's decision-making process is critical. This study assesses the performance of different pretrained Bidirectional Encoder Representations from Transformers (BERT) models and delves into understanding its decision-making within the context of medical image protocol assignment.MethodsFour different pre-trained BERT models (BERT, BioBERT, ClinicalBERT, RoBERTa) were fine-tuned for the medical image protocol classification task. Word importance was measured by attributing the classification output to every word using a gradient-based method. Subsequently, a trained radiologist reviewed the resulting word importance scores to assess the model’s decision-making process relative to human reasoning.ResultsThe BERT model came close to human performance on our test set. The BERT model successfully identified relevant words indicative of the target protocol. Analysis of important words in misclassifications revealed potential systematic errors in the model.ConclusionsThe BERT model shows promise in medical image protocol assignment by reaching near human level performance and identifying key words effectively. The detection of systematic errors paves the way for further refinements to enhance its safety and utility in clinical settings.

Comment: ChatGPT: Chasing the Storm in Radiology Training and Education.

Comment: ChatGPT: Chasing the Storm in Radiology Training and Education.

Understanding the current acute aortic syndrome (AAS) pathways-The Collaborative Acute Aortic Syndrome Project (CAASP) protocol.

Acute aortic syndrome (AAS) is an emergency associated with high peri-hospital mortality rates. Variable clinical presentation makes timely diagnosis challenging and such delays in diagnosis directly impact patient outcomes. The aims of the Collaborative Acute Aortic Syndrome Project (CAASP) are to characterise and evaluate the current AAS pathways of a cohort of hospitals in the UK, USA and New Zealand to determine if patient outcomes are influenced by the AAS pathway (time to hospital admission, diagnosis and management plan) and demographic, social, geographic and patient-specific factors (clinical presentation and comorbidities). The objectives are to describe different AAS pathways and time duration between hospital admission to diagnosis and management plan instigation, and to compare patient outcomes between pathways. The study is a multicentre, retrospective service evaluation project of adult patients diagnosed on imaging with AAS. It will be coordinated by the UK National Interventional Radiology Trainee Research (UNITE) network and Vascular and Endovascular Research Network (VERN) in conjunction with The Aortic Dissection Charitable Trust (TADCT). All AAS cases diagnosed on imaging between 1st January 2018 to 1st June 2021 will be included and followed-up for 6 months. Eligibility criteria include aortic dissection (AD) Type A, Type B, non A/B, penetrating aortic ulcer, and intramural haematoma. Exclusion criteria are non-AAS pathology, acute on chronic AAS, and age<18. This project will evaluate patient demographics, timing of presentation, patient symptoms, risk factors for AD, physical examination findings, timing to imaging and treatment, hospital stay, and mortality. Univariate and multivariate analysis will be used to identify predictors associated with prolonged time to diagnosis or treatment and mortality at 30 days.

Value of Size and Malignant Features of Lateral Lymph Nodes in Risk Stratification at Lateral Local Recurrence of Rectal Cancer: A National Cohort Study.

Patients with rectal cancer who have enlarged lateral lymph nodes (LLNs) have an increased risk of lateral local recurrence (LLR). However, little is known about prognostic implications of malignant features (internal heterogeneity, irregular margins, loss of fatty hilum, and round shape) on MRI and number of enlarged LLNs, in addition to LLN size. Of the 3,057 patients with rectal cancer included in this national, retrospective, cross-sectional cohort study, 284 with a cT3-4 tumor located ≤8 cm from the anorectal junction who received neoadjuvant treatment and who had visible LLNs on MRI were selected. Imaging was reassessed by trained radiologists. LLNs were categorized based on size. Influence of malignant features and the number of LLNs on LLR was investigated. Of 284 patients with at least 1 visible LLN, 122 (43%) had an enlarged node (≥7.0 mm) and 157 (55%) had malignant features. Of the 122 patients with enlarged nodes, 25 had multiple (≥2). In patients with a single enlarged node (n=97), a single malignant feature was associated with a 4-year LLR rate of 0% and multiple malignant features was associated with a rate of 17% (P=.060). In the group with multiple malignant features, their disappearance on restaging was associated with an LLR rate of 13% compared with an LLR rate of 20% for persistent malignant features (P=.532). The presence of intermediate-size LLNs (5.0-6.9 mm) with at least 1 malignant feature was associated with a 4-year LLR rate of 8%; the 4-year LLR rate was 13% when the malignant features persisted on restaging MRI (P=.409). Patients with multiple enlarged LLNs had a 4-year LLR rate of 28% compared with 11% for those with a single enlarged LLN (P=.059). The presence of multiple enlarged LLNs (≥7.0 mm), as well as multiple malignant features in an enlarged node contribute to the risk of developing an LLR. These radiologic features can be used for clinical decision-making regarding the potential benefit of LLN dissection.

Incorporation of Eye Tracking and Gaze Feedback to Characterize and Improve Radiologist Search Patterns of Chest X-Rays: A Randomized Controlled Clinical Trial

Diagnostic errors in radiology often result from incomplete visual assessments attributed to insufficient training in search patterns. Radiologists lack consistent feedback, relying on ad-hoc strategies, leading to suboptimal patterns and potential false negatives. This study used eye tracking to analyze search patterns, quantify performance metrics, and evaluate the impact of an automated continuous improvement feedback-driven framework on detection accuracy. In a controlled trial with ten residents, the intervention group (received feedback) demonstrated a significant 38.9% absolute improvement in detecting suspicious pulmonary nodules compared to the control group (5.6%, p=0.006). While improvements were more rapid over sessions (p=0.0001), other metrics showed no significant changes. The findings highlight the potential of automated feedback systems to enhance radiologist accuracy and reduce errors, emphasizing the need for further research and broader implementation in radiology training.

Vertebral fracture severity assessment on anteroposterior radiographs with a new semi-quantitative technique.

The aim of our study was to propose a semi-quantitative (SQ) grading scheme for osteoporotic vertebral fracture (OVF) on anteroposterior (AP) radiographs. On AP radiographs, the vertebrae are divided into right and left halves, which are graded (A) vertical rectangle, (B) square, (C) traverse rectangle, and (D) trapezoid; whole vertebrae are graded (E) transverse band or (F) bow-tie. Type A and B were compared with normal and Genant SQ grade 1 OVF, Type C and D with grade 2 OVF, and Type E and F with grade 3 OVF. Spine AP radiographs and lateral radiographs of 50 females were assessed by AP radiographs SQ grading. After training, an experienced board-certified radiologist and a radiology trainee assessed the 50 AP radiographs. The height-to-width ratio of the half vertebrae varied 1.32-1.48. On lateral radiographs, 84 vertebrae of the 50 patients had OVFs (38 grade 1, 24 grade 2, and 22 grade 3). On AP radiographs, the radiologist correctly assigned 84.2%, 91.7%, and 77.2% and the trainee correctly assigned 68.4%, 79.2%, and 81.8% of grade 1, 2, and 3 OVFs, respectively. Compared with lateral radiographs, the radiologist had a weighted Kappa of 0.944 including normal vertebrae and 0.883 not including normal vertebrae, while the corresponding Kappa values for the trainee were 0.891 and 0.830, respectively. We propose a new semi-quantitative grading system for vertebral fracture severity assessment on AP spine radiographs.

Psychological, cognitive, and physiological impact of hazards casualties' trainings on first responders: the example of a chemical and radiological training. An exploratory study.

First responders are among the first to respond to hazards casualties. They might operate in volatile, uncertain, complex, and ambiguous (VUCA) environments. While they have underlined the need to improve their knowledge and training to face these environments, there are few data regarding the stress induced by these trainings. Chemical, biological, radiological, and nuclear (CBRN) hazards casualties' trainings seem to be a good model of "in vivo" stress. First responders must operate in a hostile and encountered environment with a CBRN protective equipment that places demand on their psychological, cognitive, and physiological capacities. Current research recognizes that the activity of the parasympathetic system (PSS) can be used as an objective marker of stress adaptation, measured as heart rate variability (HRV). To compare between baseline and simulation the evolution of the parasympathetic activity (primary outcome), anxiety, emotions, cognitive load, and body posture awareness (secondary outcomes). A total of 28 first responders attended to three simulated scenarios requiring CBRN management of casualties. One day before simulation, we collected HRV data (baseline). The simulations' day (pre-, post-simulation) we collected anxiety score (STAI-Y B), emotions (SPANE), cognitive load (NASA TLX), body posture awareness (PAS) and HRV. The morning after we collected the PAS score (recovery). We compare data' evolution between different times of the simulation. (i) A high level of anxiety at baseline [Median 51 (46; 56)] which decreased between pre- and post-simulation (p = 0.04; F = 2.93); (ii) a post-simulation decrease in negative feelings (p = 0.03); (iii) a decrease in body awareness after simulation which returned to the initial level at recovery (p = 0.03; F = 3.48); (iv) a decrease in mean RR between baseline, pre- and post-simulation (p = 0.009; F = 5.11). There were no significant difference between times on others analysis of HRV. Prior to simulation, participants experienced anticipatory anxiety. Simulations training practiced regularly could be one way to combat anticipatory anxiety.

Beyond Ultrasound: Multimodal Cross-Sectional Imaging for Preoperative Imaging of Parotid Gland Tumors: A Primer for Radiology Trainees

Even if the management of parotid gland tumors depends on the histopathological subtype, preoperative imaging of parotid gland tumors is clinically relevant. Preoperative imaging gives insight into the differentiation between benign and malignant tumors, which might potentially decrease the number of unnecessary aggressive surgeries. Characteristic imaging findings on cross-sectional imaging, such as computed tomography (CT) and magnetic resonance imaging (MRI), can help narrow the differential diagnosis and guide the further management of patients presenting with parotid masses. While MRI is imperative for the determination of perineural spread, which is frequently encountered with malignant parotid tumors, CT is important for the evaluation of osseous invasion. Furthermore, multi-parametric MRI protocols provide insights into the tumor behavior and internal composition, which is helpful in the case of benign mixed tumors and others. While distant metastasis is uncommon with parotid neoplasms, PET/CT provides a valuable tool for the improved evaluation of loco-regional and distant metastatic disease. This article discusses the imaging features of common benign and malignant parotid tumors.

Hemoperitoneum Quantification in Non-contrast CT: Evaluating Feasibility with the Novel HUVAO Segmentation Algorithm

Abstract Background Injuries involving substantial bleeding, frequently encountered in victims of road traffic accidents, pose a significant risk to mortality. For abdominal trauma cases, accurately assessing internal bleeding and hematomas becomes crucial. Detecting hemoperitoneum, which indicates both blood loss and organ damage in the abdominal cavity, requires precise evaluation. Timely diagnosis and quantification of hemoperitoneum following road accidents are crucial during the critical golden hour, enabling prioritized medical intervention and potentially saving lives while enhancing overall patient care. However, achieving precise hemoperitoneum quantification in abdominal trauma faces challenges due to the intricate nature of overlapping Hounsfield unit (HU) regions. Methods In this feasibility study, we sought to assess the efficacy of the novel HUVAO (Hounsfield Unit-based Volume quantification of Asymmetrical Objects) segmentation algorithm for quantifying hemoperitoneum in thoracoabdominal non-contrast computed tomography (CT) images. Using 28 retrospective non-contrast CT scans of thoracoabdominal regions from trauma patients, we analyzed crucial imaging data without necessitating additional scans or contrast-enhanced procedures. The study aimed to compare HUVAO against classical algorithms and visual estimations by trained radiologists for hemoperitoneum segmentation in thoracoabdominal non-contrast CT images. Results Our findings revealed that although the technical feasibility of employing HUVAO and other segmentation algorithms for hemoperitoneum quantification is evident, the outcomes derived from these algorithms display notable discrepancies. Conclusion In assessing technical feasibility, we introduced the HUVAO segmentation algorithm for hemoperitoneum quantification, comparing its performance against classical segmentation algorithms and visual estimations from trained radiologists. While our results affirm the technical feasibility of HUVAO for this purpose, the observed variations underscore the task's inherent complexity. This emphasizes the limitations of relying solely on HU-based detection, advocating for integration with clinical data. This insight urges exploration of advanced techniques to boost accuracy and elevate patient care standards.