Content-based image retrieval for the diagnosis of myocardial perfusion imaging using a deep convolutional autoencoder.

BACKGROUND: Single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) plays a crucial role in the optimal treatment strategy for patients with coronary heart disease. We tested the feasibility of feature extraction from MPI using a deep convolutional autoencoder (CAE) model. METHODS: Eight hundred and forty-three pairs of stress and rest myocardial perfusion images were collected from consecutive patients who underwent cardiac scintigraphy in our hospital between December 2019 and February 2022. We trained a CAE model to reproduce the input paired image data, so as the encoder to output a 256-dimensional feature vector. The extracted feature vectors were further dimensionally reduced via principal component analysis (PCA) for data visualization. Content-based image retrieval (CBIR) was performed based on the cosine similarity of the feature vectors between the query and reference images. The agreement of the radiologist's finding between the query and retrieved MPI was evaluated using binary accuracy, precision, recall, and F1-score. RESULTS: A three-dimensional scatter plot with PCA revealed that feature vectors retained clinical information such as percent summed difference score, presence of ischemia, and the location of scar reported by radiologists. When CBIR was used as a similarity-based diagnostic tool, the binary accuracy was 81.0%. CONCLUSION: The results indicated the utility of unsupervised feature learning for CBIR in MPI.

Image similarity-based cardiac rhythm device identification from X-rays using feature point matching.

AIMS: Identifying the manufacturer and the type of cardiac implantable electronic devices (CIEDs) is important in emergent clinical settings. Recent studies have illustrated that artificial neural network models can successfully recognize CIEDs from chest X-ray images. However, all existing methods require a vast amount of medical data to train the classification model. Here, we have proposed a novel method to retrieve an identical CIED image from an image database by employing the feature point matching algorithm. METHODS AND RESULTS: A total of 653 unique X-ray images from 456 patients who visited our pacemaker clinic between April 2012 and August 2020 were collected. The device images were manually square-shaped, and was thereafter resized to 224 × 224 pixels. A scale-invariant feature transform (SIFT) algorithm was used to extract the keypoints from the query image and reference images. Paired feature points were selected via brute-force matching, and the average Euclidean distance was calculated. The image with the shortest average distance was defined as the most similar image. The classification performance was indicated by accuracy, precision, recall, and F1-score for detecting the manufacturers and model groups, respectively. The average accuracy, precision, recall, and F-1 score for the manufacturer classification were 97.0%, 0.97, 0.96, and 0.96, respectively. For the model classification task, the average accuracy, precision, recall, and F-1 score were 93.2%, 0.94, 0.92, and 0.93, respectively, all of which were higher than those of the previously reported machine learning models. CONCLUSION: Feature point matching is useful for identifying CIEDs from X-ray images.

Favorable effects of early tolvaptan administration in very elderly patients after repeat hospitalizations for acute decompensated heart failure.

Tolvaptan (TLV) is an oral selective vasopressin 2 receptor antagonist that acts on the distal nephrons, causing a loss of electrolyte-free water. To date, its early administration in very elderly patients after repeat hospitalizations for acute decompensated heart failure (ADHF) despite receiving optimal medical therapy has not been evaluated. Fifty-six ADHF patients who were >80 years old and had been repeatedly hospitalized were retrospectively enrolled in this study. Twenty-five patients (14 men; mean age 86.7 ± 5.3 years; control group) received standard therapy and 31 patients (15 men; mean age 85.5 ± 4.5 years; TLV group) received oral TLV within 24 h of admission. The rate of worsening renal function was significantly lower in the TLV group than in the control group (13 vs. 40%, P < 0.05). The duration of the return to body weight at a steady state was significantly shorter in the TLV group (5.3 ± 2.8 days) than in the control group (13.9 ± 9.2 days, P < 0.01). Consequently, the hospitalization period in the TLV group (13.5 ± 5.9 days) was significantly shorter than that in the control group (24.7 ± 12.3 days, P < 0.01). In conclusion, the early administration of TLV to very elderly patients who underwent repeat hospitalizations for ADHF resulted in immediate decongestion and thus reduced the hospitalization period with a lower incidence of worsening renal function.

Beneficial effects of adaptive servo-ventilation therapy on readmission and medical costs in patients with chronic heart failure.

Adaptive servo-ventilation (ASV) therapy is a novel modality of noninvasive positive pressure ventilation and is now widely utilized to treat patients with chronic heart failure (CHF). However, there has been no clinical study of the effect of ASV therapy on readmission and cost-effectiveness for the treatment of CHF. The present study was conducted to evaluate the clinical efficacy and cost-effectiveness of home ASV therapy in 45 patients with a history of two or more admissions a year for worsening CHF. Seven patients refused to undergo chronic ASV therapy and three died. Thus, 35 patients were eventually enrolled in the present study. New York Heart Association class (2.8 ± 0.4 versus 2.3 ± 0.5, p < 0.001), log plasma B-type natriuretic peptide level (2.53 ± 0.44 versus 2.29 ± 0.40 pg/mL, p < 0.0001), left atrial dimension (47.5 ± 7.0 versus 44.9 ± 7.6 mm, p = 0.014), and mitral regurgitation area ratio (20.3 ± 12.1 versus 16.9 ± 8.9%, p = 0.007) decreased significantly after 12 months of ASV therapy. The frequency of hospitalization after ASV was significantly lower than before ASV (1.0 ± 1.0 versus 2.3 ± 0.5 times/year/patient, p < 0.0001). ASV also decreased the duration of hospitalization from 64.4 ± 46.5 to 22.8 ± 27.5 days/year/patient (p < 0.0001). Consequently, the total medical costs were reduced by 37% after ASV (1.95 ± 1.37 versus 3.11 ± 1.75 million yen/patient, p = 0.003). ASV therapy reduced readmissions and medical costs in patients with CHF.

Left Ventricular Pseudoaneurysm With a Bead-Like Appearance After Myocardial Infarction: The Utility of Cardiac Computed Tomography.

A left ventricular pseudoaneurysm is a rare but life-threatening complication after myocardial infarction. Because untreated pseudoaneurysms have a 30%-45% risk of rupture, surgery is the preferred therapeutic option. However, its diagnosis is sometimes challenging, as a pseudoaneurysm presents with non-specific symptoms that can mimic myocardial infarction or heart failure. We report a male patient with a history of aortic dissection surgery who presented with recurrent chest pain probably due to acute coronary syndrome. Transthoracic echocardiography revealed a cavity at the apex of the left ventricle, indicating a mechanical complication after myocardial infarction. As the coronary angiography was considered difficult because of the patient's anatomical problem, contrast-enhanced computed tomography (CT) was performed. CT angiography revealed multiple nodular cavities continued from within the left ventricle. It seemed that the pseudoaneurysm was formed in stages in the adherent pericardium after myocardial infarction, resulting in a bead-like appearance. Emergent pseudoaneurysmectomy and left ventricular wall repair were performed, and the patient was discharged without any complications. This case illustrates the utility of cardiac CT to establish the diagnosis of left ventricular pseudoaneurysm and coronary artery atherosclerosis.