A deep learning-based model to estimate pulmonary function from chest x-rays: multi-institutional model development and validation study in Japan.

BACKGROUND: Chest x-ray is a basic, cost-effective, and widely available imaging method that is used for static assessments of organic diseases and anatomical abnormalities, but its ability to estimate dynamic measurements such as pulmonary function is unknown. We aimed to estimate two major pulmonary functions from chest x-rays. METHODS: In this retrospective model development and validation study, we trained, validated, and externally tested a deep learning-based artificial intelligence (AI) model to estimate forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) from chest x-rays. We included consecutively collected results of spirometry and any associated chest x-rays that had been obtained between July 1, 2003, and Dec 31, 2021, from five institutions in Japan (labelled institutions A-E). Eligible x-rays had been acquired within 14 days of spirometry and were labelled with the FVC and FEV1. X-rays from three institutions (A-C) were used for training, validation, and internal testing, with the testing dataset being independent of the training and validation datasets, and then x-rays from the two other institutions (D and E) were used for independent external testing. Performance for estimating FVC and FEV1 was evaluated by calculating the Pearson's correlation coefficient (r), intraclass correlation coefficient (ICC), mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) compared with the results of spirometry. FINDINGS: We included 141 734 x-ray and spirometry pairs from 81 902 patients from the five institutions. The training, validation, and internal test datasets included 134 307 x-rays from 75 768 patients (37 718 [50%] female, 38 050 [50%] male; mean age 56 years [SD 18]), and the external test datasets included 2137 x-rays from 1861 patients (742 [40%] female, 1119 [60%] male; mean age 65 years [SD 17]) from institution D and 5290 x-rays from 4273 patients (1972 [46%] female, 2301 [54%] male; mean age 63 years [SD 17]) from institution E. External testing for FVC yielded r values of 0·91 (99% CI 0·90-0·92) for institution D and 0·90 (0·89-0·91) for institution E, ICC of 0·91 (99% CI 0·90-0·92) and 0·89 (0·88-0·90), MSE of 0·17 L2 (99% CI 0·15-0·19) and 0·17 L2 (0·16-0·19), RMSE of 0·41 L (99% CI 0·39-0·43) and 0·41 L (0·39-0·43), and MAE of 0·31 L (99% CI 0·29-0·32) and 0·31 L (0·30-0·32). External testing for FEV1 yielded r values of 0·91 (99% CI 0·90-0·92) for institution D and 0·91 (0·90-0·91) for institution E, ICC of 0·90 (99% CI 0·89-0·91) and 0·90 (0·90-0·91), MSE of 0·13 L2 (99% CI 0·12-0·15) and 0·11 L2 (0·10-0·12), RMSE of 0·37 L (99% CI 0·35-0·38) and 0·33 L (0·32-0·35), and MAE of 0·28 L (99% CI 0·27-0·29) and 0·25 L (0·25-0·26). INTERPRETATION: This deep learning model allowed estimation of FVC and FEV1 from chest x-rays, showing high agreement with spirometry. The model offers an alternative to spirometry for assessing pulmonary function, which is especially useful for patients who are unable to undergo spirometry, and might enhance the customisation of CT imaging protocols based on insights gained from chest x-rays, improving the diagnosis and management of lung diseases. Future studies should investigate the performance of this AI model in combination with clinical information to enable more appropriate and targeted use. FUNDING: None.

Artificial intelligence-based model to classify cardiac functions from chest radiographs: a multi-institutional, retrospective model development and validation study.

BACKGROUND: Chest radiography is a common and widely available examination. Although cardiovascular structures-such as cardiac shadows and vessels-are visible on chest radiographs, the ability of these radiographs to estimate cardiac function and valvular disease is poorly understood. Using datasets from multiple institutions, we aimed to develop and validate a deep-learning model to simultaneously detect valvular disease and cardiac functions from chest radiographs. METHODS: In this model development and validation study, we trained, validated, and externally tested a deep learning-based model to classify left ventricular ejection fraction, tricuspid regurgitant velocity, mitral regurgitation, aortic stenosis, aortic regurgitation, mitral stenosis, tricuspid regurgitation, pulmonary regurgitation, and inferior vena cava dilation from chest radiographs. The chest radiographs and associated echocardiograms were collected from four institutions between April 1, 2013, and Dec 31, 2021: we used data from three sites (Osaka Metropolitan University Hospital, Osaka, Japan; Habikino Medical Center, Habikino, Japan; and Morimoto Hospital, Osaka, Japan) for training, validation, and internal testing, and data from one site (Kashiwara Municipal Hospital, Kashiwara, Japan) for external testing. We evaluated the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy. FINDINGS: We included 22 551 radiographs associated with 22 551 echocardiograms obtained from 16 946 patients. The external test dataset featured 3311 radiographs from 2617 patients with a mean age of 72 years [SD 15], of whom 49·8% were male and 50·2% were female. The AUCs, accuracy, sensitivity, and specificity for this dataset were 0·92 (95% CI 0·90-0·95), 86% (85-87), 82% (75-87), and 86% (85-88) for classifying the left ventricular ejection fraction at a 40% cutoff, 0·85 (0·83-0·87), 75% (73-76), 83% (80-87), and 73% (71-75) for classifying the tricuspid regurgitant velocity at a 2·8 m/s cutoff, 0·89 (0·86-0·92), 85% (84-86), 82% (76-87), and 85% (84-86) for classifying mitral regurgitation at the none-mild versus moderate-severe cutoff, 0·83 (0·78-0·88), 73% (71-74), 79% (69-87), and 72% (71-74) for classifying aortic stenosis, 0·83 (0·79-0·87), 68% (67-70), 88% (81-92), and 67% (66-69) for classifying aortic regurgitation, 0·86 (0·67-1·00), 90% (89-91), 83% (36-100), and 90% (89-91) for classifying mitral stenosis, 0·92 (0·89-0·94), 83% (82-85), 87% (83-91), and 83% (82-84) for classifying tricuspid regurgitation, 0·86 (0·82-0·90), 69% (68-71), 91% (84-95), and 68% (67-70) for classifying pulmonary regurgitation, and 0·85 (0·81-0·89), 86% (85-88), 73% (65-81), and 87% (86-88) for classifying inferior vena cava dilation. INTERPRETATION: The deep learning-based model can accurately classify cardiac functions and valvular heart diseases using information from digital chest radiographs. This model can classify values typically obtained from echocardiography in a fraction of the time, with low system requirements and the potential to be continuously available in areas where echocardiography specialists are scarce or absent. FUNDING: None.

Factors significantly associated with COVID-19 severity in symptomatic patients: A retrospective single-center study.

INTRODUCTION: The severity of coronavirus disease (COVID-19) in Japanese patients is unreported. We retrospectively examined significant factors associated with disease severity in symptomatic COVID-19 patients (COVID-Pts) admitted to our institution between February 20 and April 30, 2020. METHODS: All patients were diagnosed based on the genetic detection of severe acute respiratory syndrome coronavirus 2. Information on the initial symptoms, laboratory data, and computed tomography (CT) images at hospitalization were collected from the patients' records. COVID-Pts were categorized as those with critical or severe illness (Pts-CSI) or those with moderate or mild illness (Pt-MMI). All statistical analyses were performed using R software. RESULTS: Data from 61 patients (16 Pt-CSI, 45 Pt-MMI), including 58 Japanese and three East Asians, were analyzed. Pt-CSI were significantly older and had hypertension or diabetes than Pt-MMI (P < 0.001, 0.014 and < 0.001, respectively). Serum albumin levels were significantly lower in Pt-CSI than in Pt-MMI (P < 0.001), whereas the neutrophil-to-lymphocyte ratio and C-reactive protein level were significantly higher in Pt-CSI than in Pt-MMI (P < 0.001 and P < 0.001, respectively). In the CT images of 60 patients, bilateral lung lesions were more frequently observed in Pt-CSI than in Pt-MMI (P = 0.013). Among the 16 Pt-CSI, 15 received antiviral therapy, 12 received tocilizumab, five underwent methylprednisolone treatment, six received mechanical ventilation, and one died. CONCLUSIONS: The illness severity of Japanese COVID-Pts was associated with older age, hypertension and/or diabetes, low serum albumin, high neutrophil-to-lymphocyte ratio, and C-reactive protein.

Clinical and CT findings of small bowel obstruction caused by rice cakes in comparison with bezoars.

PURPOSE: Rice cakes have not been recognized as a cause of small bowel obstruction (SBO) worldwide. We compared clinical and CT findings of rice cake SBO versus SBO due to bezoars, the most common cause of food-induced SBO. METHODS: Twenty-four patients with rice cake SBO (n = 17) or bezoar SBO (n = 7) were retrospectively evaluated for clinical findings and the following multi-detector CT (MDCT) features: identification of the transition zone, presence of intraluminal lesions, degree of obstruction, and length and attenuation of obstructing materials. Categorical variables were compared by Fisher's exact test, and continuous variables by independent t test. RESULTS: None of the rice cake SBO patients required surgery, whereas 4/7 (57%) bezoar SBO patients underwent surgery. On MDCT, rice cake residues were recognized as well-defined intraluminal lesions of shorter length (29.8 ± 4.6 mm vs. 47.7 ± 10.8 mm for bezoars; p < 0.0001) and higher attenuation (106 ± 27.8 HU vs. - 62.8 ± 14.7 HU for bezoars; p < 0.0001). CONCLUSIONS: Rice cake SBO patients did not require surgery. On MDCT, rice cake residues were significantly shorter and higher in attenuation than bezoars. These findings facilitate diagnosis and support the conservative management of rice cake SBO.

Cystic lesions of the parotid gland: radiologic-pathologic correlation according to the latest World Health Organization 2017 Classification of Head and Neck Tumours.

The latest World Health Organization 2017 Classification of Head and Neck Tumours includes a new chapter on tumors and tumor-like lesions of the neck and lymph nodes. Tumor-like lesions include a variety of cystic lesions of the parotid gland. Cystic lesions of the parotid gland can be divided into three groups: non-neoplastic cysts, benign tumors with macrocystic change, and malignant tumors with macrocystic change. It is important to distinguish these lesions from one another because treatment and patient management differ among the three groups. The purpose of this review is to describe the magnetic resonance imaging and scintigraphy findings and the histopathologic characteristics of each parotid gland lesion based on the latest World Health Organization 2017 Classification of Head and Neck Tumours and to summarize the key points of differential diagnosis for cystic lesions of the parotid gland.

Internal Hernias in the Era of Multidetector CT: Correlation of Imaging and Surgical Findings.

Clinical diagnosis of internal hernias is challenging because of their nonspecific signs and symptoms. Many types of internal hernias have been defined: paraduodenal, small bowel mesentery-related, greater omentum-related, lesser sac, transverse mesocolon-related, pericecal, sigmoid mesocolon-related, falciform ligament, pelvic internal, and Roux-en-Y anastomosis-related. An internal hernia is a surgical emergency that can develop into intestinal strangulation and ischemia. Accurate preoperative diagnosis is crucial for appropriate management. Multidetector computed tomography (CT), with its thin-section axial images, high-quality multiplanar reformations, and three-dimensional images, currently plays an essential role in preoperative diagnosis of internal hernias. The diagnostic approach to internal hernias at multidetector CT includes detecting an intestinal closed loop, identifying the hernia orifice, and analyzing abnormal displacement of surrounding structures and key vessels around the hernia orifice and hernia sac. At each step, multidetector CT can depict pathognomonic findings. A saclike appearance suggests an intestinal closed loop in several types of internal hernias. Convergence, engorgement, and twisting of mesenteric vessels in the hernia orifice can be seen clearly at multidetector CT, especially with use of multiplanar reformations. For definitive diagnosis of an internal hernia, analysis of displacement of anatomic landmarks around the hernia orifice is particularly important, and thin-section images provide the required information. Detailed knowledge of the anatomy, etiology, and imaging landmarks of the various hernia types is also necessary. Familiarity with the appearances of internal hernias at multidetector CT allows accurate and specific preoperative diagnosis. (©)RSNA, 2015.

Diffusion-weighted magnetic resonance imaging findings in a patient with struma ovarii.

In this report, the magnetic resonance imaging (MRI) appearance of struma ovarii (SO) in a patient who underwent diffusion-weighted imaging (DWI) of the pelvis and subsequent histological analysis is described. The solid portion of SO showed a high apparent diffusion coefficient (ADC) value, indicating unrestricted diffusion, and each loculus of SO showed different ADC values due to the different viscosity of the cyst contents in each loculus. These unique and characteristic DWI findings may serve as a helpful sign in making the correct diagnosis of SO when DWI findings are interpreted in conjunction with conventional MRI findings.

Factors associated with aggravation of esophageal varices after B-RTO for gastric varices.

PURPOSE: To retrospectively evaluate risk factors for aggravation of esophageal varices (EV) within 1 year after balloon-occluded retrograde transvenous obliteration (B-RTO) of gastric varices (GV) and to clarify suitable timing for upper endoscopy to detect EV aggravation after B-RTO. METHODS: Participants included 67 patients who underwent B-RTO for GV between January 2006 and December 2010. Whether EV aggravation occurred within 1 year was evaluated, and the time interval from B-RTO to aggravation was calculated. Factors potentially associated with EV aggravation were analyzed. RESULTS: B-RTO was successfully performed in all patients. EV aggravation at 1 year after B-RTO was found in 38 patients (56.7 %). Multivariate logistic regression analysis showed that total bilirubin (T-bil) (P = 0.032) and hepatic venous pressure gradient (HVPG) (P = 0.011) were significant independent risk factors for EV aggravation after B-RTO. Cutoff values of T-bil and HVPG yielding maximal combined sensitivity and specificity for EV aggravation were 1.6 mg/dL and 13 mmHg, respectively. The patients with T-bil ≥ 1.6 mg/dL or HVPG ≥ 13 mmHg had a median aggravation time of 5.1 months. All five patients with ruptured EV belonged to this group. In contrast, patients with T-bil < 1.6 mg/dL and HVPG < 13 mmHg had a median aggravation time of 21 months. CONCLUSION: T-bil and HVPG were significant independent risk factors for EV aggravation after B-RTO. The patients with T-bil ≥ 1.6 mg/dL or HVPG ≥ 13 mmHg require careful follow-up evaluation, including endoscopy.

Small bowel obstruction due to an intramesosigmoid hernia diagnosed by multidetector row computed tomography: a case report.

Small bowel obstruction caused by an internal hernia involving the sigmoid mesocolon is rare, and this condition is difficult to diagnose clinically. We herein report a case of small bowel obstruction due to an intramesosigmoid hernia that was diagnosed by a 64-row multidetector computed tomography and surgically treated.

CNS complications of hematopoietic stem cell transplantation.

OBJECTIVE: With the worldwide increase in the use of hematopoietic stem cell transplantation (HSCT), a high level of diligence is required for radiologists to understand HSCT-related complications in the CNS. This article describes the clinical background of HSCT and complications that occur in a time-dependent manner through the course of HSCT and addresses pivotal issues in diagnostic imaging. CONCLUSION: Acknowledging the realm of imaging manifestations and the underlying mechanism of HSCT will enhance diagnostic accuracy and optimize treatment decisions.

Pituitary abscess with panhypopituitarism showing T1 signal hyperintensity of the marginal pituitary area: a non-invasive differential diagnosis of pituitary abscess and pituitary apoplexy.

A 53-year-old man was hospitalized with general fatigue, headache, dizziness and polyuria. The laboratory findings revealed anterior hypopituitarism and central diabetes insipidus. He also showed eye movement disorder and facial sensory impairment. These symptoms were treated successfully with conservative medical treatment. Concurrently, abnormal pituitary MR imaging findings were revealed. Pituitary abscess was primarily suspected on MR imaging findings, although it was difficult to differentiate pituitary apoplexy by MR imaging findings, alone. In this report, we propose a new diagnostic approach of pituitary abscess, using a combination of CT, MR imaging and clinical manifestations, without either pituitary surgery or pituitary biopsy.

Diffusion-weighted magnetic resonance imaging in tubo-ovarian abscess: a case report.

In this report, we describe the magnetic resonance imaging appearance of tubo-ovarian abscess (TOA) in a patient who underwent diffusion-weighted imaging (DWI) of the pelvis and subsequent histologic analysis. The content of the TOA was markedly high signal intensity relative to the surrounding tissue on DWI, and it showed low signal intensity on the apparent diffusion coefficient map; these findings were consistent with those of published reports on brain and liver abscesses.

Brown tumor with fluid-fluid levels in a patient with primary hyperparathyroidism: radiological findings.

We report a case of a brown tumor with fluid-fluid levels in a patient with primary hyperparathyroidism. A 19-year-old woman presented with a 3-month history of pain in the left pubic region. The laboratory data showed elevated serum calcium and intact parathyroid hormone, confirming the diagnosis of primary hyperparathyroidism. Plain radiography and computed tomography (CT) showed an expansile lytic lesion of the superior ramus of the left pubis. The cortex was thinned. On magnetic resonance (MR) images, the lesion was solid and cystic. The solid area of the lesion showed heterogeneous low to intermediate signal intensity on T1-weighted images and heterogeneous low to high signal intensity on T2-weighted images. The cystic area showed several fluid-fluid levels on T2-weighted images. Dynamic contrast-enhanced MR images after administration of Gd-DTPA showed marked, early enhancement of the solid area of the lesion. A delayed image showed prolonged enhancement of the solid area and enhancement of the septa and walls of the cystic area. Histopathology of a biopsy specimen showed fibroblastic proliferation, abundant giant cells, and focal hemosiderin deposition, which supported the diagnosis of a brown tumor. After removing the parathyroid adenoma, the brown tumor regressed and became sclerotic on radiographs.

Radiosensitive functional dye: clinical application for estimation of patient skin dose.

Institutional review board approval and informed patient consent were obtained. The purpose of the study was to prospectively evaluate the use of radiosensitive indicators to estimate patient entrance skin dose (ESD). Forty-six patients wore a jacket with 48 or 52 indicators adhered to the back during percutaneous coronary interventions; they had eight additional indicators on their upper arms. The patients' ESDs were calculated according to the change in color of the indicators. There were good correlations between the ESDs estimated by using color measurements performed with an optical instrument and those estimated at visual observation (P < .001) and between the ESDs estimated by using a thermoluminescent dosimeter and those estimated by using color measurements (P < .001). The radiosensitive indicator method seems to be useful for estimating ESDs and their distribution during percutaneous coronary intervention; however, visual observation is reliable for estimating doses of up to 5 Gy only. (c) RSNA, 2006.

Regional cerebral blood flow in a patient with Nasu-Hakola disease.

We report a functional neuroimaging study of a 43-year-old woman with Nasu-Hakola disease (NHD). Regional cerebral blood flow (rCBF) images were measured with technetium-99m ethyl cysteinate dimer single photon emission computed tomography (SPECT). rCBF was decreased in the bilateral frontal lobes and thalamus. This finding was consistent with the known underlying neuropathology in patients with NHD. Brain SPECT is useful for demonstrating the pathophysiologic brain region in patients with NHD.

Noncardiogenic pulmonary edema as the first manifestation of pheochromocytoma: a case report.

We report here a case of pheochromocytoma presenting with noncardiogenic pulmonary edema, which is an unusual first manifestation of pheochromocytoma. Chest radiograph showed diffuse consolitaion and ground-glass opacity on both lungs. Neither pleural effusion nor cardiomegaly was present. High-resolution computed tomography (HRCT) showed widespread ground-glass opacity with fine intralobular reticular opacity (so-called "crazy-paving" appearance) and consolidation in a strikingly dependent distribution. The areas of ground-glass opacity and consolidation were sharply marginated by interlobular septa, showing a geographic appearance. The pulmonary edema and dyspnea resolved over three days without cardioactive or diuretic treatment. Noncardiogenic pulmonary edema is an unusual first manifestation of pheochromocytoma. It is important to recognize this unusual manifestation in the emergency situation.

Ovarian fibroma (fibrothecoma) with extensive cystic degeneration: unusual MR imaging findings in two cases.

Two cases of ovarian fibroma (fibrothecoma) with extensive cystic degeneration are described. One case revealed a large multiloculated cystic mass with a small solid part. Each locule showed variable signal intensity on both T1- and T2-weighted magnetic resonance (MR) images. The other case revealed a parviloculated cystic mass with a small solid part. In both cases, the solid parts were located at the periphery of the mass and showed distinct low signal intensity on both T1- and T2-weighted MR images and slight enhancement. It should be noted that ovarian fibromas (fibrothecomas) with extensive cystic degeneration are a rare exception to the rule that solid components in a cystic adnexal mass imply malignancy.

Benign pleomorphic adenoma with extensive cystic degeneration: unusual MR findings in two cases.

Two cases of benign pleomorphic adenoma (PA) with extensive cystic degeneration are described. In both cases magnetic resonance (MR) imaging revealed entirely cystic tumors with an irregularity at the cyst wall. Some additional findings were observed: one case showed layering within the cyst contents caused by intratumoral hemorrhage, and the other had a small mural nodule at the cyst wall that showed very low intensity on both T1- and T2-weighted images and corresponded pathologically to the hyaline stroma. In both cases, preoperative diagnosis was difficult based on the MR findings alone. Benign PA should be included in the differential diagnosis of an entirely cystic parotid mass.