A Machine Learning Approach for Knee Injury Detection from Magnetic Resonance Imaging.

The knee is an essential part of our body, and identifying its injuries is crucial since it can significantly affect quality of life. To date, the preferred way of evaluating knee injuries is through magnetic resonance imaging (MRI), which is an effective imaging technique that accurately identifies injuries. The issue with this method is that the high amount of detail that comes with MRIs is challenging to interpret and time consuming for radiologists to analyze. The issue becomes even more concerning when radiologists are required to analyze a significant number of MRIs in a short period. For this purpose, automated tools may become helpful to radiologists assisting them in the evaluation of these images. Machine learning methods, in being able to extract meaningful information from data, such as images or any other type of data, are promising for modeling the complex patterns of knee MRI and relating it to its interpretation. In this study, using a real-life imaging protocol, a machine-learning model based on convolutional neural networks used for detecting medial meniscus tears, bone marrow edema, and general abnormalities on knee MRI exams is presented. Furthermore, the model's effectiveness in terms of accuracy, sensitivity, and specificity is evaluated. Based on this evaluation protocol, the explored models reach a maximum accuracy of 83.7%, a maximum sensitivity of 82.2%, and a maximum specificity of 87.99% for meniscus tears. For bone marrow edema, a maximum accuracy of 81.3%, a maximum sensitivity of 93.3%, and a maximum specificity of 78.6% is reached. Finally, for general abnormalities, the explored models reach 83.7%, 90.0% and 84.2% of maximum accuracy, sensitivity and specificity, respectively.

Patello-Femoral Pain Syndrome: Magnetic Resonance Imaging versus Ultrasound.

BACKGROUND: Magnetic Resonance Imaging (MRI) and Ultrasound (US) in combination with clinical data could contribute to the diagnosis, staging and follow-up of Patello-Femoral Syndrome (PFS), which often overlaps with other pathologies of the knee. PURPOSE OF THE STUDY: To evaluate the diagnostic role of MRI and US findings associated with PFS and define the range values of instrumental measurements obtained in pathological cases and healthy controls, the performance of the two methods in comparison, and the correlation with clinical data. MATERIALS AND METHODS: 100 subjects were examined: 60 patients with a high suspicion of PFS at the clinical evaluation and 40 healthy controls. All measurements obtained by MRI and US examination were correlated with clinical data. A descriptive analysis of all measurements was stratified for pathological cases and healthy controls. A Student's t-test for continuous variables was used to compare patients to controls and US to MRI. Logistic regression analysis was applied to test the correlation between MRI and US measurements with clinical data. RESULTS: Statistical descriptive analysis determined the MRI and US range values of medial patello-femoral distance and the thickness of retinacles and cartilages in pathological cases and healthy controls. In pathological cases, the retinacle results of both increased; the medial appeared to be slightly more increased than the lateral. Furthermore, in some cases, the thickness of the cartilage decreased in both techniques; the medial cartilage was more thinned than the lateral. According to logistic regression analyses, the best diagnostic parameter was the medial patello-femoral distance due to the overlapping results of the US and MRI. Furthermore, all clinical data obtained by different tests showed a good correlation with patello-femoral distance. In particular, the correlation between medial patello-femoral distance and the VAS score is direct and equal to 97-99%, which is statistically significant (p < 0.001), and the correlation with the KOOS score is inverse and equal to 96-98%, which is statistically significant. CONCLUSIONS: MRI and Ultrasound examination in combination with clinical data demonstrated high-value results in the diagnosis of PFS.

Myocardial iron overload assessed by magnetic resonance imaging (MRI)T2* in multi-transfused patients with thalassemia and acquired anemias.

BACKGROUND: Cardiac complications secondary to iron overload remain a significant matter in patients with transfusion dependent anemias. PATIENTS AND METHODS: To evaluate cardiac siderosis, Magnetic resonance imaging T2* (MRI T2*) was performed in 3 cohorts of transfusion dependent patients: 99 with thalassemia major (TM), 20 with thalassemia intermedia (TI), and 10 with acquired anemias (AA). Serum ferritin was measured and all patients underwent echocardiographic evaluation. RESULTS: In TM patients cardiac T2* pathologic values (below 20 ms) were found in 37 patients. Serum ferritin was negatively associated with age (r=-0.32, p=0.001) and weakly with T2* values (r=-0.19, p=0.057). A positive correlation was found between T2* and LVEF (r=0.27, p=0.006). Out of 37 patients with T2*<20 ms, 18 (48%) had serum ferritin values<1000 ng/ml. In TI cohort, 3 patients had cardiac T2* pathologic values. In AA cohort, pathologic T2* values were found in 2 patients, who received 234 and 199 PRBC units, respectively, and were both on chelation therapy (in one patient ferritin value was 399 ng/ml). T2* values were negatively associated, but not significantly, with the number of PRBC transfused (r=-0.53, p=0.07). CONCLUSION: In our experience, 37% of TM patients had a myocardial iron overload assessed by MRI T2*; this value is higher than in TI patients. Serum ferritin measurement was a poor predictor of myocardial siderosis. In patients with AA, more than 200 PRBC units transfused were required to induce cardiac hemosiderosis, in spite of chelation therapy and, in one patient, of normal ferritin values.