-
PloS One 2021The chest X-ray (CXR) is the most readily available and common imaging modality for the assessment of pneumonia. However, detecting pneumonia from chest radiography is a...
OBJECTIVE
The chest X-ray (CXR) is the most readily available and common imaging modality for the assessment of pneumonia. However, detecting pneumonia from chest radiography is a challenging task, even for experienced radiologists. An artificial intelligence (AI) model might help to diagnose pneumonia from CXR more quickly and accurately. We aim to develop an AI model for pneumonia from CXR images and to evaluate diagnostic performance with external dataset.
METHODS
To train the pneumonia model, a total of 157,016 CXR images from the National Institutes of Health (NIH) and the Korean National Tuberculosis Association (KNTA) were used (normal vs. pneumonia = 120,722 vs.36,294). An ensemble model of two neural networks with DenseNet classifies each CXR image into pneumonia or not. To test the accuracy of the models, a separate external dataset of pneumonia CXR images (n = 212) from a tertiary university hospital (Gachon University Gil Medical Center GUGMC, Incheon, South Korea) was used; the diagnosis of pneumonia was based on both the chest CT findings and clinical information, and the performance evaluated using the area under the receiver operating characteristic curve (AUC). Moreover, we tested the change of the AI probability score for pneumonia using the follow-up CXR images (7 days after the diagnosis of pneumonia, n = 100).
RESULTS
When the probability scores of the models that have a threshold of 0.5 for pneumonia, two models (models 1 and 4) having different pre-processing parameters on the histogram equalization distribution showed best AUC performances of 0.973 and 0.960, respectively. As expected, the ensemble model of these two models performed better than each of the classification models with 0.983 AUC. Furthermore, the AI probability score change for pneumonia showed a significant difference between improved cases and aggravated cases (Δ = -0.06 ± 0.14 vs. 0.06 ± 0.09, for 85 improved cases and 15 aggravated cases, respectively, P = 0.001) for CXR taken as a 7-day follow-up.
CONCLUSIONS
The ensemble model combined two different classification models for pneumonia that performed at 0.983 AUC for an external test dataset from a completely different data source. Furthermore, AI probability scores showed significant changes between cases of different clinical prognosis, which suggest the possibility of increased efficiency and performance of the CXR reading at the diagnosis and follow-up evaluation for pneumonia.
Topics: Adult; Aged; Area Under Curve; Artificial Intelligence; Female; Humans; Male; Middle Aged; Pneumonia; ROC Curve; Tertiary Care Centers; Thorax; Tomography, X-Ray Computed
PubMed: 33857181
DOI: 10.1371/journal.pone.0249399 -
Clinical Oral Investigations Mar 2021Chronic rhinosinusitis (CRS) frequently stems from a dental origin, although odontogenic sinusitis (OS) remains underdiagnosed amongst different professionals. This...
OBJECTIVES
Chronic rhinosinusitis (CRS) frequently stems from a dental origin, although odontogenic sinusitis (OS) remains underdiagnosed amongst different professionals. This study aimed to explore how often odontogenic causes are considered when diagnosing CRS.
MATERIALS AND METHODS
Patient records from 374 new CRS patients treated at a tertiary-level ear, nose, and throat (ENT) clinic were selected. Entries and radiological reports were assessed exploring how often dentition was mentioned and OS was suspected, how often radiologists reported maxillary teeth, and how commonly typical OS microbial findings and unilateral symptoms occurred.
RESULTS
Although 10.1% of the CRS diagnoses were connected to possible dental issues, teeth were not mentioned for 73.8% of patients. Radiological reports were available from 267 computed or cone beam computed tomographies, of which 25.1% did not mention the maxillary teeth. The reported maxillary teeth pathology was not considered in 31/64 (48.4%) cases. Unilateral symptoms associated with apical periodontitis (OR = 2.49, 95% CI 1.27-4.89, p = 0.008). Microbial samples were available from 88 patients, for whom Staphylococcus aureus was the most common finding (17% of samples).
CONCLUSIONS
Odontogenic causes are often overlooked when diagnosing CRS. To provide adequate treatment, routine assessment of patient's dental history and status, careful radiograph evaluation, and utilization of microbial findings should be performed. Close cooperation with dentists is mandatory.
CLINICAL RELEVANCE
Dental professionals should be aware of difficulties medical professionals encounter when diagnosing possible OS. Thus, sufficient knowledge of OS pathology is essential to both medical and dental professionals.
Topics: Cone-Beam Computed Tomography; Humans; Maxillary Sinusitis; Odontogenesis; Sinusitis; Tomography, X-Ray Computed
PubMed: 32500403
DOI: 10.1007/s00784-020-03384-4 -
Malaria Journal Jul 2022Despite the large-scale rollout of malaria rapid diagnostic tests (RDTs) in Tanzania, many healthcare providers (HCPs) continue using blood film microscopy (BFM) and...
Knowledge, attitudes and practices towards malaria diagnostics among healthcare providers and healthcare-seekers in Kondoa district, Tanzania: a multi-methodological situation analysis.
BACKGROUND
Despite the large-scale rollout of malaria rapid diagnostic tests (RDTs) in Tanzania, many healthcare providers (HCPs) continue using blood film microscopy (BFM) and clinical examination to diagnose malaria, which can increase the risk of mal-diagnosis and over-prescribing of anti-malarials. Patients disregarding negative test results and self-treating exacerbate the problem. This study explored the knowledge, attitudes and practices of HCPs and healthcare-seekers regarding RDTs in comparison to BFM testing.
METHODS
A situational analysis was, therefore, conducted in Kondoa District, Dodoma Region, Tanzania. A multi-methodological approach was adopted including (i) a health facility inventory and screening of logbooks from May 2013 to April 2014 with 77,126 patient entries from 33 health facilities; (ii) a survey of 40 HCPs offering malaria services; and iii) a survey of 309 randomly selected household members from the facilities' catchment area. Surveys took place in April and May 2014.
RESULTS
Health facility records revealed that out of 77,126 patient entries, 22% (n = 17,235) obtained a malaria diagnosis. Of those, 45% were made with BFM, 33% with RDT and 22% with clinical diagnosis. A higher rate of positive diagnoses was observed with BFM compared with RDT (71% vs 14%). In the HCP survey, 48% preferred using BFM for malaria testing, while 52% preferred RDT. Faced with a negative RDT result for a patient presenting with symptoms typical for malaria, 25% of HCPs stated they would confirm the result with a microscopy test, 70% would advise or perform a clinical diagnosis and 18% would prescribe anti-malarials. Interviews with household members revealed a preference for microscopy testing (58%) over RDT (23%), if presented with malaria symptoms. For participants familiar with both tests, a second opinion was desired in 45% after a negative microscopy result and in 90% after an RDT.
CONCLUSIONS
Non-adherence to negative diagnostics by HCPs and patients continues to be a concern. Frequent training and supportive supervision for HCPs diagnosing and treating malaria and non-malaria febrile illnesses is essential to offer quality services that can instil confidence in HCPs and patients alike. The introduction of new diagnostic devices should be paired with context-specific behaviour change interventions targeting healthcare-seekers and healthcare providers.
Topics: Antimalarials; Diagnostic Tests, Routine; Health Facilities; Health Knowledge, Attitudes, Practice; Health Personnel; Humans; Malaria; Tanzania
PubMed: 35864543
DOI: 10.1186/s12936-022-04244-0 -
Arquivos de Gastroenterologia 2021Non-alcoholic fatty liver disease (NAFLD) is the most prevalent liver disease in the world, and its prevalence is increasing alongside obesity. In United States, NAFLD...
BACKGROUND
Non-alcoholic fatty liver disease (NAFLD) is the most prevalent liver disease in the world, and its prevalence is increasing alongside obesity. In United States, NAFLD is already the second leading cause of liver transplantation. The spectrum of the disease ranges from simple steatosis, which has a benign course, to steatohepatitis, which may progress to cirrhosis and its complications. The rising of noninvasive methods for diagnosing and staging non-alcoholic steatohepatitis (NASH) and fibrosis decreases the need of liver biopsy, as well as the costs and the occurrence of complications related to it.
OBJECTIVE
To analyze the performance of the triglyceride-glucose index to evaluate steatosis, NASH and liver fibrosis in obese patients with NAFLD.
METHODS
This is a retrospective cross-sectional study. Every medical record of patients who were candidates for bariatric surgery at a leading hospital in Southern Brazil were analyzed. The triglyceride-glucose index (TyG Index), a method composed only of two simple laboratory tests (serum triglycerides and fasting glucose levels), was performed prior to surgery. The TyG Index performance regarding the anatomopathological findings was evaluated, and the AUROC curve was calculated to evaluate the best cut-off point for diagnosing steatosis, non-alcoholic steatohepatitis and liver fibrosis grade. Also, the NAFLD fibrosis Score (NFS) was evaluated.
RESULTS
A total of 423 patients were evaluated. The TyG Index with a cut-off point of 8.76 excluded significant simple steatosis (grade 2-3) in obese patients, with 67.6% sensitivity, 65.1% specificity, 46.3% positive predictive value (PPV), 81.8% negative predictive value (NPV), 65.8% accuracy and 0.66 AUROC (P=0.005). In the evaluation of NASH, the TyG Index with a cut-off point of 8.82 excluded significant NASH (grade 2-3) with 57.3% sensitivity, 58.6% specificity, 33.7% PPV, 78.8% NPV, 58.2% accuracy and 0.58 AUROC (P=0.022). When evaluating liver fibrosis, the TyG Index with a cut-off point of 8.91 showed a sensitivity of 61.8%, a specificity of 62.5%, a PPV of 13.8 and a NPV of 94.4% for exclusion of advanced fibrosis (F3-4), with a 62.4% accuracy and 0.69 AUROC (P<0.001). When analyzing the performance of NFS in the diagnosis of advanced fibrosis, the cut-off point <-1.455 excluded advanced fibrosis with sensitivity of 59.4%, specificity of 51%, PPV of 11%, NPV of 92.4% and accuracy of 51.7%. However, the cut-off point of 0.676 to diagnose advanced fibrosis presented sensitivity of 21.9%, specificity of 83%, PPV of 11.7%, NPV of 91.2% and 77.3% accuracy. The AUROC was 0.54 (P=0.480).
CONCLUSION
TyG Index did not perform well in the diagnosis of significant steatosis and NASH. However, it was able to exclude advanced fibrosis in obese patients who are candidates for bariatric surgery.
Topics: Biopsy; Cross-Sectional Studies; Glucose; Humans; Liver; Liver Cirrhosis; Non-alcoholic Fatty Liver Disease; Obesity; Retrospective Studies; Triglycerides
PubMed: 34231659
DOI: 10.1590/S0004-2803.202100000-24 -
Acta Clinica Croatica Apr 2023Lymph node biopsy is indicated in patients with suspected malignancy or lymphadenopathy due to unclarified reasons. Lymph node biopsy can be performed as fine needle...
Lymph node biopsy is indicated in patients with suspected malignancy or lymphadenopathy due to unclarified reasons. Lymph node biopsy can be performed as fine needle aspiration biopsy, core biopsy, or excisional lymph node biopsy. In particular, the diagnosis of malignant lymphoma is considered insufficient for oncological treatment unless classified into subgroups. Core biopsy and excisional biopsy can be performed to diagnose lymphoma and classify it into subgroups. Core biopsy may also be limited in some cases for the diagnosis of lymphoma. Therefore, patients are referred to surgical departments for excisional lymph node biopsy. It was aimed herein to analyze the results of excisional lymph node biopsies performed for diagnostic purposes in our department. Data on 73 patients having undergone diagnostic excisional lymph node biopsy at Sakarya University Medical Faculty Training and Research Hospital between January 2008 and January 2020 were retrospectively analyzed. Patients were evaluated in terms of age, gender, biopsy site, pathological diagnosis, number and diameter of lymph nodes excised. Patients younger than 18 years of age, those with sentinel lymph node biopsies, and lymph node dissections performed for any known malignancy were excluded from the study. Statistical data analysis was done using SPSS statistical software. There were 37 (50.7%) female and 36 (49.3%) male patients, mean age 52.07 (18-90) years. Axillary lymph node biopsy was performed in 32 patients, inguinal lymph node biopsy in 29 patients, cervical lymph node biopsy in 3 patients, intra-abdominal lymph node biopsy in 6 patients, mediastinal lymph node biopsy in 1 patient, and supraclavicular lymph node biopsy in 2 patients. All of the lymph node biopsies were performed as excisional biopsy. Malignancy was detected in 36 (49.3%) patients. In 37 (50.3%) patients, the causes of lymphadenopathy were found to be benign pathologies. When the causes of malignant disease were examined, it was observed that 23 (31.5%) patients were diagnosed with lymphoma. Hodgkin lymphoma was detected in 5 patients diagnosed with lymphoma, and non-Hodgkin lymphoma was found in 18 patients. Metastatic lymphadenopathy was observed in 13 (17.8%) patients. Reactive lymphoid hyperplasia (26%) and lymphadenitis (20.5%) were found among the causes of benign lymphadenopathy. The number of excised lymph nodes was between 1 and 4, and their diameter was between 9 and 75 mm (mean: 29.53±15.56 mm). There was no statistically significant difference between benign and malignant patients according to gender, age, lymph node diameter, number of lymph nodes excised, and excisional lymph node biopsy site. For diagnostic lymph node biopsy, fine-needle aspiration biopsy and core biopsy should be performed primarily. If lymphoma is suspected in the diagnosis, fine-needle aspiration biopsy is not necessary. In this case, it is believed that it is more appropriate to perform core biopsy first. If the core biopsy is insufficient for diagnosis, it is more appropriate to perform surgical biopsy in order to cause no delay in diagnosis and treatment. Excisional biopsy is a method that can be safely performed and does not cause severe morbidity in palpable peripheral lymphadenopathies. Although it does not cause severe morbidity because it is an invasive procedure, excisional biopsy should be performed in a selected patient group.
Topics: Humans; Male; Female; Middle Aged; Retrospective Studies; Lymph Nodes; Biopsy; Lymphadenopathy; Lymphoma; Hodgkin Disease; Biopsy, Fine-Needle
PubMed: 38304357
DOI: 10.20471/acc.2023.62.01.07 -
Endoscopy International Open Oct 2023Pancreatic cysts are common incidental findings, with an estimated prevalence of 13% to 15% in imaging done for other reasons. Diagnosis often relies on collection of...
Pancreatic cysts are common incidental findings, with an estimated prevalence of 13% to 15% in imaging done for other reasons. Diagnosis often relies on collection of cyst fluid, but tissue sampling using micro-forceps may allow for a more reliable diagnosis and higher yield of DNA for next-generation sequencing (NGS). The primary aim was to assess the performance of NGS in identifying mucinous cyst. The secondary aims were to assess DNA yield between the cyst fluid and cyst wall tissue, complication rate and performance of conventional investigations. Twenty-four patients referred for endoscopic ultrasound were recruited. Biopsies were taken using micro-forceps and the AmpliSeq Cancer Hotspot panel was used for NGS, a polymerase chain reaction assay targeting several hotspots within 50 genes, including , and . The concentration of DNA extracted from 24 cyst wall samples was significantly higher than in the nine of 24 available matched cyst fluid samples. The sensitivity, specificity, and diagnostic accuracy of NGS for diagnosing mucinous cyst were 93%, 50% and 84%; for standard of care, they were -66.6%, 50% and 63.1%; and for standard of care with NGS, they were 100%, 50%, and 89.4% respectively. Cyst wall biopsy was able to diagnose 19 of 24 cysts (4 high risk, 7 intraductal papillary mucinous neoplasms, 4 cysts of mucinous origin, and 4 benign). NGS data correlate well with histology and may aid in diagnosis and risk stratification of pancreatic cysts. Cyst wall biopsy performs well in diagnosing cysts but was inadequate in five of 24 patients.
PubMed: 37941539
DOI: 10.1055/a-2163-8805 -
Medicine Nov 2023Transverse testicular ectopia (TTE) is a rare congenital malformation with a high rate of misdiagnosis and mistreatment before operation, which cannot be diagnosed even... (Review)
Review
RATIONALE
Transverse testicular ectopia (TTE) is a rare congenital malformation with a high rate of misdiagnosis and mistreatment before operation, which cannot be diagnosed even during operation due to lack of knowledge.
PATIENT CONCERNS
Two patients with ectopic testes who were misdiagnosed as right inguinal hernia for the first time and underwent surgery. The "ovary" and "testicle" like structures was seen in the right inguinal region during the first operation. After being transferred to our hospital for laparoscopic surgery, it was found that the left spermatic cord and testis were transversely transverted to the right, the left testis was fixed at the right inner ring, and agglomerated soft tissue could be seen in the right inguinal canal, which was suspected to be Muller tube.
DIAGNOSES
Based on preoperative images and intraoperative findings, both cases were diagnosed with Transverse testicular ectopia (TTE). The postoperative pathology report for the second patient revealed the presence of an in situ spermatogenic cell tumor in the ectopic testis.
INTERVENTIONS
Preperitoneal tension-free repair of right inguinal hernia and resection of left cryptorchidism were performed on the 2 patients.
OUTCOMES
Postoperative pathology of the first patient confirmed that the resected specimens contained tubal-like and uterine-like structures. The postoperative pathology of the second patient showed that the resected tissue consists of immature testis, epididymis, uterus and seminal vesicle glands, in which an in situ spermatogenic tumor could be seen in the testicular tissue. Postoperative diagnosis: left transversal testicular ectopia and right indirect inguinal hernia.
LESSONS
The clinical misdiagnosis and mistreatment rate of TTE is very high. Once the patients with cryptorchidism complicated with inguinal hernia are found in clinic, the possibility of the disease must be considered. For the patients whose cryptorchidism does not descend into the ipsilateral scrotum and it is difficult to diagnose, laparoscopy can be used for both diagnosis and treatment. If a patient has both inguinal hernia and cryptorchidism, it is crucial to rule out a diagnosis of TTE to prevent misdiagnosis and inappropriate treatment.
Topics: Humans; Male; Cryptorchidism; Diagnostic Errors; Hernia, Inguinal; Neoplasms; Testis
PubMed: 37933018
DOI: 10.1097/MD.0000000000035850 -
BMC Endocrine Disorders Jun 2023To compare the ability of the Cox regression and machine learning algorithms to predict the survival of patients with Anaplastic thyroid carcinoma (ATC).
BACKGROUND
To compare the ability of the Cox regression and machine learning algorithms to predict the survival of patients with Anaplastic thyroid carcinoma (ATC).
METHODS
Patients diagnosed with ATC were extracted from the Surveillance, Epidemiology, and End Results database. The outcomes were overall survival (OS) and cancer-specific survival (CSS), divided into: (1) binary data: survival or not at 6 months and 1 year; (2): time-to-event data. The Cox regression method and machine learnings were used to construct models. Model performance was evaluated using the concordance index (C-index), brier score and calibration curves. The SHapley Additive exPlanations (SHAP) method was deployed to interpret the results of machine learning models.
RESULTS
For binary outcomes, the Logistic algorithm performed best in the prediction of 6-month OS, 12-month OS, 6-month CSS, and 12-month CSS (C-index = 0.790, 0.811, 0.775, 0.768). For time-event outcomes, traditional Cox regression exhibited good performances (OS: C-index = 0.713; CSS: C-index = 0.712). The DeepSurv algorithm performed the best in the training set (OS: C-index = 0.945; CSS: C-index = 0.834) but performs poorly in the verification set (OS: C-index = 0.658; CSS: C-index = 0.676). The brier score and calibration curve showed favorable consistency between the predicted and actual survival. The SHAP values was deployed to explain the best machine learning prediction model.
CONCLUSIONS
Cox regression and machine learning models combined with the SHAP method can predict the prognosis of ATC patients in clinical practice. However, due to the small sample size and lack of external validation, our findings should be interpreted with caution.
Topics: Humans; Thyroid Carcinoma, Anaplastic; Algorithms; Databases, Factual; Machine Learning; Thyroid Neoplasms; Prognosis
PubMed: 37291551
DOI: 10.1186/s12902-023-01368-5 -
Scientific Reports Mar 2021To study the usefulness of virtual reality (VR)-based training for diagnosing strabismus. Fourteen residents in ophthalmology performed at least 30 VR training sessions...
To study the usefulness of virtual reality (VR)-based training for diagnosing strabismus. Fourteen residents in ophthalmology performed at least 30 VR training sessions to diagnose esotropia and exotropia. Examinations of real patients with esotropia or exotropia before and after the VR training were video-recorded and presented to a strabismus expert to assess accuracy and performance scores for measuring the deviation angle and diagnosing strabismus with anonymization. A feedback survey regarding the usefulness and ease of use of the VR application was conducted for participants. The mean age of the 14 ophthalmology residents (10 men and 4 women), was 29.7 years. Before VR training, participants showed a mean accuracy score of 14.50 ± 5.45 and a performance score of 9.64 ± 4.67 for measuring the deviation angle and diagnosing strabismus in real patients with strabismus. After VR training, they showed a significantly improved accuracy score of 22.14 ± 4.37 (p = 0.012) and a performance score of 15.50 ± 1.99 (p = 0.011). According to the survey, most participants agreed on the usefulness of VR applications. This study suggests that VR-based training improved ophthalmology residents' clinical diagnostic skills for strabismus in a short period.
Topics: Adult; Educational Measurement; Female; Humans; Male; Strabismus; Surveys and Questionnaires; Virtual Reality
PubMed: 33723316
DOI: 10.1038/s41598-021-85265-8 -
Brazilian Journal of Otorhinolaryngology 2021Hearing impairment, or hearing loss, can be caused by several factors and the implications vary according to the type, degree, cause and age of occurrence. Hearing...
INTRODUCTION
Hearing impairment, or hearing loss, can be caused by several factors and the implications vary according to the type, degree, cause and age of occurrence. Hearing screening should be a common procedure, allowing the pre-clinical identification and necessary referrals, avoiding the consequences of auditory deprivation. Mobile apps have shown to be a good alternative for hearing assessment.
OBJECTIVE
The objective was to develop an app and assess its performance in identifying hearing loss by comparing it with another validated screening tool.
METHODS
The application, called Ouviu, was created using audiological knowledge and tools available on the iOS platform. We evaluated 185 people, aged between 6 and 96 years, distributed into 5 age groups, performing audiometry and hearing screening using two tools: HearCheck and the application.
RESULTS
The results showed that the sensitivity of the application in identifying hearing loss was approximately 97%, while that of HearCheck was 79%. The positive predictive value of the application showed that the probability of a person being identified with this tool and actually having a hearing loss was 94%, while for HearCheck it was 96%. False negatives, which failed hearing loss identification, were fewer in the app (3%) than in HearCheck (21%).
CONCLUSION
Consequently, the developed application was shown to be effective as a hearing screening tool, surpassing HearCheck in identifying mild hearing loss. In addition to being portable, easy to apply, low cost and rapidly performed, the application has the advantage of assessing environmental noise to perform the exam, as well as the fact that it is not necessary to attach any hardware to the mobile device.
Topics: Adolescent; Adult; Aged; Aged, 80 and over; Audiometry; Child; Hearing; Hearing Loss; Humans; Mass Screening; Middle Aged; Mobile Applications; Young Adult
PubMed: 32620319
DOI: 10.1016/j.bjorl.2020.03.009