-
International Journal of Surgery... Jun 2024Colorectal cancer (CRC) stands as the third most prevalent cancer globally, projecting 3.2 million new cases and 1.6 million deaths by 2040. Accurate lymph node... (Meta-Analysis)
Meta-Analysis
BACKGROUND
Colorectal cancer (CRC) stands as the third most prevalent cancer globally, projecting 3.2 million new cases and 1.6 million deaths by 2040. Accurate lymph node metastasis (LNM) detection is critical for determining optimal surgical approaches, including preoperative neoadjuvant chemoradiotherapy and surgery, which significantly influence CRC prognosis. However, conventional imaging lacks adequate precision, prompting exploration into radiomics, which addresses this shortfall by converting medical images into reproducible, quantitative data.
METHODS
Following PRISMA, Supplemental Digital Content 1 (http://links.lww.com/JS9/C77) and Supplemental Digital Content 2 (http://links.lww.com/JS9/C78), and AMSTAR-2 guidelines, Supplemental Digital Content 3 (http://links.lww.com/JS9/C79), we systematically searched PubMed, Web of Science, Embase, Cochrane Library, and Google Scholar databases until 11 January 2024, to evaluate radiomics models' diagnostic precision in predicting preoperative LNM in CRC patients. The quality and bias risk of the included studies were assessed using the Radiomics Quality Score (RQS) and the modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Subsequently, statistical analyses were conducted.
RESULTS
Thirty-six studies encompassing 8039 patients were included, with a significant concentration in 2022-2023 (20/36). Radiomics models predicting LNM demonstrated a pooled area under the curve (AUC) of 0.814 (95% CI: 0.78-0.85), featuring sensitivity and specificity of 0.77 (95% CI: 0.69, 0.84) and 0.73 (95% CI: 0.67, 0.78), respectively. Subgroup analyses revealed similar AUCs for CT and MRI-based models, and rectal cancer models outperformed colon and colorectal cancers. Additionally, studies utilizing cross-validation, 2D segmentation, internal validation, manual segmentation, prospective design, and single-center populations tended to have higher AUCs. However, these differences were not statistically significant. Radiologists collectively achieved a pooled AUC of 0.659 (95% CI: 0.627, 0.691), significantly differing from the performance of radiomics models (P<0.001).
CONCLUSION
Artificial intelligence-based radiomics shows promise in preoperative lymph node staging for CRC, exhibiting significant predictive performance. These findings support the integration of radiomics into clinical practice to enhance preoperative strategies in CRC management.
Topics: Humans; Colorectal Neoplasms; Lymphatic Metastasis; Lymph Nodes; Radiomics
PubMed: 38935817
DOI: 10.1097/JS9.0000000000001239 -
The Indian Journal of Radiology &... Jul 2024Although abundant literature is currently available on the use of deep learning for breast cancer detection in mammography, the quality of such literature is widely... (Review)
Review
Although abundant literature is currently available on the use of deep learning for breast cancer detection in mammography, the quality of such literature is widely variable. To evaluate published literature on breast cancer detection in mammography for reproducibility and to ascertain best practices for model design. The PubMed and Scopus databases were searched to identify records that described the use of deep learning to detect lesions or classify images into cancer or noncancer. A modification of Quality Assessment of Diagnostic Accuracy Studies (mQUADAS-2) tool was developed for this review and was applied to the included studies. Results of reported studies (area under curve [AUC] of receiver operator curve [ROC] curve, sensitivity, specificity) were recorded. A total of 12,123 records were screened, of which 107 fit the inclusion criteria. Training and test datasets, key idea behind model architecture, and results were recorded for these studies. Based on mQUADAS-2 assessment, 103 studies had high risk of bias due to nonrepresentative patient selection. Four studies were of adequate quality, of which three trained their own model, and one used a commercial network. Ensemble models were used in two of these. Common strategies used for model training included patch classifiers, image classification networks (ResNet in 67%), and object detection networks (RetinaNet in 67%). The highest reported AUC was 0.927 ± 0.008 on a screening dataset, while it reached 0.945 (0.919-0.968) on an enriched subset. Higher values of AUC (0.955) and specificity (98.5%) were reached when combined radiologist and Artificial Intelligence readings were used than either of them alone. None of the studies provided explainability beyond localization accuracy. None of the studies have studied interaction between AI and radiologist in a real world setting. While deep learning holds much promise in mammography interpretation, evaluation in a reproducible clinical setting and explainable networks are the need of the hour.
PubMed: 38912238
DOI: 10.1055/s-0043-1775737 -
Diagnostics (Basel, Switzerland) May 2024This study delves into the cutting-edge field of deep learning techniques, particularly deep convolutional neural networks (DCNNs), which have demonstrated unprecedented... (Review)
Review
OBJECTIVES
This study delves into the cutting-edge field of deep learning techniques, particularly deep convolutional neural networks (DCNNs), which have demonstrated unprecedented potential in assisting radiologists and orthopedic surgeons in precisely identifying meniscal tears. This research aims to evaluate the effectiveness of deep learning models in recognizing, localizing, describing, and categorizing meniscal tears in magnetic resonance images (MRIs).
MATERIALS AND METHODS
This systematic review was rigorously conducted, strictly following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Extensive searches were conducted on MEDLINE (PubMed), Web of Science, Cochrane Library, and Google Scholar. All identified articles underwent a comprehensive risk of bias analysis. Predictive performance values were either extracted or calculated for quantitative analysis, including sensitivity and specificity. The meta-analysis was performed for all prediction models that identified the presence and location of meniscus tears.
RESULTS
This study's findings underscore that a range of deep learning models exhibit robust performance in detecting and classifying meniscal tears, in one case surpassing the expertise of musculoskeletal radiologists. Most studies in this review concentrated on identifying tears in the medial or lateral meniscus and even precisely locating tears-whether in the anterior or posterior horn-with exceptional accuracy, as demonstrated by AUC values ranging from 0.83 to 0.94.
CONCLUSIONS
Based on these findings, deep learning models have showcased significant potential in analyzing knee MR images by learning intricate details within images. They offer precise outcomes across diverse tasks, including segmenting specific anatomical structures and identifying pathological regions. Contributions: This study focused exclusively on DL models for identifying and localizing meniscus tears. It presents a meta-analysis that includes eight studies for detecting the presence of a torn meniscus and a meta-analysis of three studies with low heterogeneity that localize and classify the menisci. Another novelty is the analysis of arthroscopic surgery as ground truth. The quality of the studies was assessed against the CLAIM checklist, and the risk of bias was determined using the QUADAS-2 tool.
PubMed: 38893617
DOI: 10.3390/diagnostics14111090 -
European Journal of Surgical Oncology :... Jun 2024Confidence in long-term treatment results of radiofrequency ablation (RFA) for papillary thyroid microcarcinoma (PTMC) is required in comparison with surgery and active... (Review)
Review
OBJECTIVES
Confidence in long-term treatment results of radiofrequency ablation (RFA) for papillary thyroid microcarcinoma (PTMC) is required in comparison with surgery and active surveillance (AS). The objective of this meta-analysis is to report more than three years of follow-up results of radiofrequency ablation for PTMCs.
METHODS
Ovid PUBMED, COCHRANE, and EMBASE databases were searched through Nov 19, 2023, for studies reporting outcomes in patients with PTMC treated with radiofrequency ablation and followed up for more than 3 years. The standard mean difference of the tumor volume before and after therapy, tumor recurrence, lymph node (LN) metastasis, distant metastasis, complications, and the pooled volume reduction rates (VRRs) at 1, 3, 6, 12, 24, 36, and 48 months after radiofrequency ablation were assessed. Data were extracted and methodological quality was assessed independently by two radiologists according to the PRISMA guidelines.
RESULTS
Eight studies, involving 2131 patients, met the inclusion criteria through database searches. The overall VRR was 99.81 % (95 % CI: 99.68, 99.95) in the last follow-up. During a mean pooled follow-up of 46.59 months, 69 patients experienced local PTMC recurrence, with 8 cases within the ablation area. Additionally, 44 patients were diagnosed with newly discovered PTMC, and 17 patients exhibited lymph node metastases. Among the patients with PTMC recurrence, 3 were under active surveillance while 59 underwent additional RFA. The pooled mean complication rate was 2.80 %, with no instances of life-threatening or delayed complications.
CONCLUSIONS
Radiofrequency ablation proves to be an effective local tumor control method for low-risk PTMC patients, resulting in clinically significant and enduring volume reduction. The rate of regrowth and retreatment requirement post-RFA was notably lower, positioning RFA as a compelling alternative to existing treatment options.
PubMed: 38870871
DOI: 10.1016/j.ejso.2024.108470 -
Annals of Surgery Jun 2024To systematically review technologies that objectively measure CWL in surgery, assessing their psychometric and methodological characteristics.
OBJECTIVE
To systematically review technologies that objectively measure CWL in surgery, assessing their psychometric and methodological characteristics.
SUMMARY BACKGROUND DATA
Surgical tasks involving concurrent clinical decision-making and the safe application of technical and non-technical skills require a substantial cognitive demand and resource utilization. Cognitive overload leads to impaired clinical decision-making and performance decline. Assessing cognitive workload (CWL) could enable interventions to alleviate burden and improve patient safety.
METHODS
Ovid MEDLINE, OVID Embase, the Cochrane Library and IEEE Xplore databases were searched from inception to August 2023. Full-text, peer-reviewed original studies in a population of surgeons, anesthesiologists or interventional radiologists were considered, with no publication date constraints. Study population, task paradigm, stressor, Cognitive Load Theory (CLT) domain, objective and subjective parameters, statistical analysis and results were extracted. Studies were assessed for a) definition of CWL, b) details of the clinical task paradigm, and c) objective CWL assessment tool. Assessment tools were evaluated using psychometric and methodological characteristics.
RESULTS
10790 studies were identified; 9004 were screened; 269 full studies were assessed for eligibility, of which 67 met inclusion criteria. The most widely used assessment modalities were autonomic (32 eye studies and 24 cardiac). Intrinsic workload (e.g. task complexity) and germane workload (effect of training or expertize) were the most prevalent designs investigated. CWL was not defined in 30 of 67 studies (44.8%). Sensitivity was greatest for neurophysiological instruments (100% EEG, 80% fNIRS); and across modalities accuracy increased with multi-sensor recordings. Specificity was limited to cardiac and ocular metrics, and was found to be sub-optimal (50% and 66.67%). Cardiac sensors were the least intrusive, with 54.2% of studies conducted in naturalistic clinical environments (higher ecological validity).
CONCLUSION
Physiological metrics provide an accessible, objective assessment of CWL, but dependence on autonomic function negates selectivity and diagnosticity. Neurophysiological measures demonstrate favorable sensitivity, directly measuring brain activation as a correlate of cognitive state. Lacking an objective gold standard at present, we recommend the concurrent use of multimodal objective sensors and subjective tools for cross-validation. A theoretical and technical framework for objective assessment of CWL is required to overcome the heterogeneity of methodological reporting, data processing, and analysis.
PubMed: 38847099
DOI: 10.1097/SLA.0000000000006370 -
European Urology Oncology May 2024Prostate Imaging for Recurrence Reporting (PI-RR) was introduced in 2021 to standardize the interpretation and reporting of multiparametric magnetic resonance imaging... (Review)
Review
Prostate Magnetic Resonance Imaging Using the Prostate Imaging for Recurrence Reporting (PI-RR) Scoring System to Detect Recurrent Prostate Cancer: A Systematic Review and Meta-analysis.
BACKGROUND AND OBJECTIVE
Prostate Imaging for Recurrence Reporting (PI-RR) was introduced in 2021 to standardize the interpretation and reporting of multiparametric magnetic resonance imaging (MRI) for prostate cancer following whole-gland treatment. The system scores image on a scale from 1 to 5 and has shown promising results in single-center studies. The aim of our systematic review and meta-analysis was to assess the diagnostic performance of the PI-RR system in predicting the likelihood of local recurrence after whole-gland treatment.
METHODS
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for diagnostic test accuracy were followed. Relevant databases were searched up to December 2023. Primary studies met the eligibility criteria if they reported MRI diagnostic performance in prostate cancer recurrence using PI-RR. Diagnostic performance for MRI was assessed using two different cutoff points (≥3 or ≥4 for positivity according to the PI-RR system). A meta-analysis with a random-effects model was used to estimate pooled sensitivity and specificity values.
KEY FINDINGS AND LIMITATIONS
Sixteen articles were identified for full-text reading, of which six were considered eligible, involving a total of 467 patients. Using a cutoff of PI-RR ≥3 (4 studies) for recurrent disease, the sensitivity was 77.8% (95% confidence interval [CI] 69.9-84.1%) and the specificity was 80.2% (95% CI 58.2-92.2%). Using a cutoff of PI-RR ≥4 (4 studies), the sensitivity was 61.9% (95% CI 35.6-82.7%) and the specificity was 86.6% (95% CI 75.1-93.3%). Overall, the inter-rater agreement varied from fair to excellent.
CONCLUSIONS AND CLINICAL IMPLICATIONS
PI-RR is accurate in detecting local recurrence after whole-gland treatment for prostate cancer and shows fair-to-good to excellent inter-reader agreement. Overall, a PI-RR cutoff of ≥3 showed high sensitivity and specificity.
PATIENT SUMMARY
We reviewed studies that reported on how good MRI scans using a scoring system called PI-RR were in detecting recurrence of prostate cancer. We found that this system shows good performance, with fair to excellent agreement between different radiologists.
PubMed: 38824004
DOI: 10.1016/j.euo.2024.05.007 -
Radiology. Cardiothoracic Imaging Jun 2024Purpose To perform a systematic review and meta-analysis to assess the prognostic value of stress perfusion cardiac MRI in predicting cardiovascular outcomes. Materials... (Meta-Analysis)
Meta-Analysis
Prognostic Value of Stress Perfusion Cardiac MRI in Cardiovascular Disease: A Systematic Review and Meta-Analysis of the Effects of the Scanner, Stress Agent, and Analysis Technique.
Purpose To perform a systematic review and meta-analysis to assess the prognostic value of stress perfusion cardiac MRI in predicting cardiovascular outcomes. Materials and Methods A systematic literature search from the inception of PubMed, Embase, Web of Science, and China National Knowledge Infrastructure until January 2023 was performed for articles that reported the prognosis of stress perfusion cardiac MRI in predicting cardiovascular outcomes. The quality of included studies was assessed using the Quality in Prognosis Studies tool. Reported hazard ratios (HRs) of univariable regression analyses with 95% CIs were pooled. Comparisons were performed across different analysis techniques (qualitative, semiquantitative, and fully quantitative), magnetic field strengths (1.5 T vs 3 T), and stress agents (dobutamine, adenosine, and dipyridamole). Results Thirty-eight studies with 58 774 patients with a mean follow-up time of 53 months were included. There were 1.9 all-cause deaths and 3.5 major adverse cardiovascular events (MACE) per 100 patient-years. Stress-inducible ischemia was associated with a higher risk of all-cause mortality (HR: 2.55 [95% CI: 1.89, 3.43]) and MACE (HR: 3.90 [95% CI: 2.69, 5.66]). For MACE, pooled HRs of qualitative, semiquantitative, and fully quantitative methods were 4.56 (95% CI: 2.88, 7.22), 3.22 (95% CI: 1.60, 6.48), and 1.78 (95% CI: 1.39, 2.28), respectively. For all-cause mortality, there was no evidence of a difference between qualitative and fully quantitative methods ( = .79). Abnormal stress perfusion cardiac MRI findings remained prognostic when subgrouped based on underlying disease, stress agent, and field strength, with HRs of 3.54, 2.20, and 3.38, respectively, for all-cause mortality and 3.98, 3.56, and 4.21, respectively, for MACE. There was no evidence of subgroup differences in prognosis between field strengths or stress agents. There was significant heterogeneity in effect size for MACE outcomes in the subgroups assessing qualitative versus quantitative stress perfusion analysis, underlying disease, and field strength. Conclusion Stress perfusion cardiac MRI is valuable for predicting cardiovascular outcomes, regardless of the analysis method, stress agent, or magnetic field strength used. MR-Perfusion, MRI, Cardiac, Meta-Analysis, Stress Perfusion, Cardiac MR, Cardiovascular Disease, Prognosis, Quantitative © RSNA, 2024
Topics: Humans; Prognosis; Cardiovascular Diseases; Magnetic Resonance Imaging; Myocardial Perfusion Imaging; Exercise Test
PubMed: 38814186
DOI: 10.1148/ryct.230382 -
Abdominal Radiology (New York) May 2024This review aims to provide a comprehensive summary of DECT techniques, acquisition workflows, and post-processing methods. By doing so, we aim to elucidate the...
PURPOSE
This review aims to provide a comprehensive summary of DECT techniques, acquisition workflows, and post-processing methods. By doing so, we aim to elucidate the advantages and disadvantages of DECT compared to conventional single-energy CT imaging.
METHODS
A systematic search was conducted on MEDLINE/EMBASE for DECT studies in liver imaging published between 1980 and 2024. Information regarding study design and endpoints, patient characteristics, DECT technical parameters, radiation dose, iodinated contrast agent (ICA) administration and postprocessing methods were extracted. Technical parameters, including DECT phase, field of view, pitch, collimation, rotation time, arterial phase timing (from injection), and venous timing (from injection) from the included studies were reported, along with formal narrative synthesis of main DECT applications for liver imaging.
RESULTS
Out of the initially identified 234 articles, 153 met the inclusion criteria. Extensive variability in acquisition parameters was observed, except for tube voltage (80/140 kVp combination reported in 50% of articles) and ICA administration (1.5 mL/kg at 3-4 mL/s, reported in 91% of articles). Radiation dose information was provided in only 40% of articles (range: 6-80 mGy), and virtual non-contrast imaging (VNC) emerged as a common strategy to reduce the radiation dose. The primary application of DECT post-processed images was in detecting focal liver lesions (47% of articles), with predominance of study focusing on hepatocellular carcinoma (HCC) (27%). Furthermore, a significant proportion of the articles (16%) focused on enhancing DECT protocols, while 15% explored metastasis detection.
CONCLUSION
Our review recommends using 80/140 kVp tube voltage with 1.5 mL/kg ICA at 3-4 mL/s flow rate. Post-processing should include low keV-VMI for enhanced lesion detection, IMs for tumor iodine content evaluation, and VNC for dose reduction. However, heterogeneous literature hinders protocol standardization.
PubMed: 38811447
DOI: 10.1007/s00261-024-04380-y -
Endocrine May 2024To evaluate the safety and efficacy of radiofrequency ablation (RFA) in treating locoregional recurrent thyroid cancer (LRTC) after a 2-year follow-up time.
PURPOSE
To evaluate the safety and efficacy of radiofrequency ablation (RFA) in treating locoregional recurrent thyroid cancer (LRTC) after a 2-year follow-up time.
METHODS
PubMed, Embase and Cochrane Library were searched from inception until 20 September 2022 to find studies reporting the safety and efficacy of RFA in LRTC patients after a 2-year follow-up. Two radiologists performed the data extraction and methodological quality assessment according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.
RESULTS
We analyzed 6 studies, 229 LRTC patients with 319 locally recurrent tumors were treated with RFA. The mean follow-up time of each study was ≥24 months. The pooled changes in the largest diameter and volume were 7.22 mm (95% confidence interval (CI), 6.35-8.09 mm) and 164.28 mm (95% CI, 87.78-240.77 mm), respectively; the pooled volume reduction rate was 95.03% (95% CI, 87.56-102.49%). The total complete disappearance rate after treatment was 92% (95% CI, 83-100%). The pooled decrease of serum thyroglobulin levels was 0.02 ng/ml (95% CI, -0.00-0.04 ng/ml). The pooled proportion of recurrence rate was 6% (95% CI, 0-13%). The pooled complication rate was 5% (95% CI, 0-10%). The major complications were voice change and hoarseness, only one patient developed permanent vocal cord paralysis; minor complications were cough and pain.
CONCLUSIONS
Ultrasound-guided RFA is an effective and safe treatment for LRTC based on 2-year follow-up results.
PubMed: 38801598
DOI: 10.1007/s12020-023-03660-9 -
Cureus Apr 2024This review article explores spinal injuries in athletes participating in various sporting activities. It also highlights the various mechanisms of injuries that... (Review)
Review
This review article explores spinal injuries in athletes participating in various sporting activities. It also highlights the various mechanisms of injuries that contribute to spinal injuries in each sport. Electronic databases such as PubMed, Cochrane Library, Web of Science, Embase, MEDLINE Ovid, and Google Scholar were searched for articles from 2000 to 2022 on spine injuries in sports and radiological studies discussing the various injury patterns among athletes. Studies were scoured in accordance with the inclusion criteria, and relevant data such as the number of participants, sporting activities, spine injuries, and outcomes were retrieved. Fifteen articles that met the inclusion criteria were included in the study. Cervical spine injuries are common in athletes who participate in contact sports such as football. Similarly, athletes in collision sports such as football, rugby, and hockey are likely to suffer stingers due to traction and compression injuries. Players engaged in such as soccer, baseball, and swimming, are likely to suffer from spondylolysis. Soccer players are more prone to multiple lesions compared to athletes in sports such as baseball because the sport involves training exercises such as jogging and running without kicking any ball. In swimmers, spondylolysis is common in breaststroke and butterfly styles since they involve repeated flexion and hyperextension of the lumbar spine. CT is essential for diagnosing spondylolysis as it demonstrates the lesions more accurately. Ice hockey is associated with a significant incidence of cervical spine injuries, mostly due to players being constantly checked/pushed from behind. Spine injuries are common in elite athletes across several sports. About 10% of spinal injuries in the United States result from sports activities. In diagnosing spine injuries, imaging modalities such as MRI, CT, or plain radiographs are essential. From a radiologist's perspective, these tests help immensely in deciding which treatment is required for a particular athlete or how the injury can be optimally managed. Achieving recovery from a specific spine injury usually depends on the kind of injury and the rehabilitation process the athletes undergo before returning to play.
PubMed: 38784300
DOI: 10.7759/cureus.58780