Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 307.292
Filtrar
Mais filtros








Intervalo de ano de publicação
1.
Swiss Dent J ; 133(2): 89-96, 2023 Feb 06.
Artigo em Francês | MEDLINE | ID: mdl-36723432

RESUMO

When taking intraoral X-rays with a round collimator, the salivary glands and the thyroid gland are exposed to ionizing radiation. The rectangular collimator reduces the patient's radiation dose. This study aims to analyze the impact of the rectangular collimator on the quality of intraoral X-rays taken by students of the University Clinics of Dental Medicine (Geneva, Switzerland). Intraoral X-rays, performed by 8 students with digital intraoral phosphor plates, were collected and a quality score was assigned to each X-ray. A total of 130 intraoral radiographs were analyzed. Half of the radiographs were taken with a rectangular collimator and 24.6% (32/130) were bitewings. The mean quality score of bitewings taken with or without a collimator was 11 ± 1.8 and 10.3 ± 0.9, respectively. The mean quality score of apical X-rays taken with or without collimator was 11.6 ± 1.0 and 11.5 ± 2.1, respectively. X-rays taken with a rectangular collimator were of excellent or sufficient diagnostic quality for 95% of the bitewings and 100% of the periapical radiographs. This quality control study showing that students perform quality intraoral X-rays with a rectangular collimator could encourage clinicians to use this device as ameans of radiation protection in their daily practice.


Assuntos
Radiografia Dentária Digital , Radiografia Dentária , Humanos , Doses de Radiação , Radiografia , Pescoço
2.
Sci Rep ; 13(1): 1815, 2023 Feb 01.
Artigo em Inglês | MEDLINE | ID: mdl-36725901

RESUMO

Hallux valgus (HV) is a common foot deformity that is more prevalent in females, characterised by abnormal adduction of the first metatarsal (MT) and valgus deviation of the phalanx on the transverse plane. Increasing evidence indicates that HV is more than a 2D deformity but a 3D one with rotational malalignment. Pronation deformity is seen during clinical examination for HV patients, but the exact origin of this rotational deformity is still unknown. Some attribute it to first tarsometatarsal (TMT) joint rotation, while others attribute it to intra-metatarsal bony torsion. In addition, the correlation between the rotational and transverse plane deformity is inconclusive. Identifying the origin of the rotational deformity will help surgeons choose the optimal surgical procedure while also enhancing our understanding of the pathophysiology of HV. This study aims to (1) develop an objective method for measuring the first MT torsion and first TMT joint rotation; (2) investigate the exact location of the coronal deformity in HV; (3) investigate the relationship between the severity of deformity on the transverse and coronal planes as well as the correlation between deformity severity and foot function/symptoms in HV. Age-matched females with and without HV were recruited at the Foot and Ankle Clinic of the Department of Orthopaedics and Traumatology. Computed tomography was conducted for all subjects with additional weight-bearing dorsal-plantar X-ray examination for HV subjects. Demographic information of all subjects was recorded, with symptoms and functions related to HV evaluated. The intra-class correlation was used to explore the relationship between deformities on different planes and the deformity severity and functional outcomes, respectively. An Independent t-test was used to compare joint rotation and bone torsion degrees. TMT joint rotation is significantly correlated with foot function. HV patients had more TMT joint rotation but not MT torsion compared to normal controls. No relationship was found between the coronal rotation and the 1,2-intermetatarsal angle (IMA) or Hallux valgus angle (HVA) on the transverse plane. Our results indicate that coronal deformities in HV may originate from TMT joint rotation. In addition, the severity of the TMT joint coronal rotation correlates with worse foot function; thus, multi-plane assessment and examination will be necessary for more precise surgical correction.


Assuntos
Hallux Valgus , Articulação Metatarsofalângica , Feminino , Humanos , Hallux Valgus/diagnóstico por imagem , Hallux Valgus/cirurgia , Radiografia , Tomografia Computadorizada por Raios X , Articulação Metatarsofalângica/cirurgia , Osteotomia/métodos , Estudos Retrospectivos
3.
Rofo ; 195(2): 160, 2023 Feb.
Artigo em Alemão | MEDLINE | ID: mdl-36724773

Assuntos
Radiologia , Radiografia
5.
Nature ; 614(7946): 45-47, 2023 Feb.
Artigo em Inglês | MEDLINE | ID: mdl-36725992

RESUMO

Ultra-stripped supernovae are different from other terminal explosions of massive stars, as they show little or no ejecta from the actual supernova event1,2. They are thought to occur in massive binary systems after the exploding star has lost its surface through interactions with its companion2. Such supernovae produce little to no kick, leading to the formation of a neutron star without loss of the binary companion, which itself may also evolve into another neutron star2. Here we show that a recently discovered high-mass X-ray binary, CPD -29 2176 (CD -29 5159; SGR 0755-2933)3-6, has an evolutionary history that shows the neutron star component formed during an ultra-stripped supernova. The binary has orbital elements that are similar both in period and in eccentricity to 1 of 14 Be X-ray binaries that have known orbital periods and eccentricities7. The identification of the progenitors systems for ultra-stripped supernovae is necessary as their evolution pathways lead to the formation of binary neutron star systems. Binary neutron stars, such as the system that produced the kilonova GW170817 that was observed with both electromagnetic and gravitational energy8, are known to produce a large quantity of heavy elements9,10.


Assuntos
Gravitação , Astros Celestes , Raios X , Radiografia , Nêutrons
6.
Sci Rep ; 13(1): 887, 2023 Jan 17.
Artigo em Inglês | MEDLINE | ID: mdl-36650270

RESUMO

Chest X-rays are the most economically viable diagnostic imaging test for active pulmonary tuberculosis screening despite the high sensitivity and low specificity when interpreted by clinicians or radiologists. Computer aided detection (CAD) algorithms, especially convolution based deep learning architecture, have been proposed to facilitate the automation of radiography imaging modalities. Deep learning algorithms have found success in classifying various abnormalities in lung using chest X-ray. We fine-tuned, validated and tested EfficientNetB4 architecture and utilized the transfer learning methodology for multilabel approach to detect lung zone wise and image wise manifestations of active pulmonary tuberculosis using chest X-ray. We used Area Under Receiver Operating Characteristic (AUC), sensitivity and specificity along with 95% confidence interval as model evaluation metrics. We also utilized the visualisation capabilities of convolutional neural networks (CNN), Gradient-weighted Class Activation Mapping (Grad-CAM) as post-hoc attention method to investigate the model and visualisation of Tuberculosis abnormalities and discuss them from radiological perspectives. EfficientNetB4 trained network achieved remarkable AUC, sensitivity and specificity of various pulmonary tuberculosis manifestations in intramural test set and external test set from different geographical region. The grad-CAM visualisations and their ability to localize the abnormalities can aid the clinicians at primary care settings for screening and triaging of tuberculosis where resources are constrained or overburdened.


Assuntos
Aprendizado Profundo , Tuberculose Pulmonar , Tuberculose , Humanos , Raios X , Radiografia , Tuberculose Pulmonar/diagnóstico por imagem , Pulmão/diagnóstico por imagem
7.
Artigo em Inglês | MEDLINE | ID: mdl-36674348

RESUMO

The rise of artificial intelligence (AI) in medicine, and particularly in radiology, is becoming increasingly prominent. Its impact will transform the way the specialty is practiced and the current and future education model. The aim of this study is to analyze the perception that undergraduate medical students have about the current situation of AI in medicine, especially in radiology. A survey with 17 items was distributed to medical students between 3 January to 31 March 2022. Two hundred and eighty-one students correctly responded the questionnaire; 79.3% of them claimed that they knew what AI is. However, their objective knowledge about AI was low but acceptable. Only 24.9% would choose radiology as a specialty, and only 40% of them as one of their first three options. The applications of this technology were valued positively by most students, who give it an important Support Role, without fear that the radiologist will be replaced by AI (79.7%). The majority (95.7%) agreed with the need to implement well-established ethical principles in AI, and 80% valued academic training in AI positively. Surveyed medical students have a basic understanding of AI and perceive it as a useful tool that will transform radiology.


Assuntos
Radiologia , Estudantes de Medicina , Humanos , Inteligência Artificial , Radiologia/educação , Radiografia , Inquéritos e Questionários
10.
Radiography (Lond) ; 29(1): 227-233, 2023 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-36608376

RESUMO

INTRODUCTION: Understanding the current ICT-related experience is essential for planning and effectively implementing quality healthcare services. Hence, this study aims to assess the knowledge and utilisation of ICT among radiographers in Sri Lanka. METHODS: A descriptive cross-sectional study was conducted among 590 practicing radiographers in Sri Lanka. Data was collected through a postal survey using a structured self-administered questionnaire. The questionnaire consisted of three sections: socio-demographic characteristics, existing knowledge of ICT, and utilisation of ICT applications and facilities. RESULTS: A total of 416 radiographers returned the questionnaire giving a response rate of 70.5%. Considering the overall ICT knowledge, 24.0% of the respondents possessed good knowledge, while 54.3% and 21.6% reported having fair and poor knowledge, respectively. The knowledge of ICT was significantly associated with gender, age, level of education, duration of service, and previous ICT training experience (p < 0.05). Digital radiography and electronic patient record (EPR) systems were used by 8% and 17.8% of respondents, respectively. Inadequate ICT facilities (56.7%) were identified as the most significant challenge for radiographers to use ICT. CONCLUSIONS: The majority of the respondents in this study had a fair knowledge of ICT, and this knowledge was significantly associated with certain demographic factors. Further, it was found that access to certain ICT applications, such as digital radiography and EPR systems, is limited. Hence, this study highlighted the importance of providing systematic, comprehensive and regular ICT training programmes and improving access to ICT facilities for radiographers. IMPLICATIONS OF PRACTICE: The study provides insight into the significance of improving ICT literacy among radiographers in the field. In addition, the findings may draw policymakers' attention to improving radiographers' access to the latest technologies.


Assuntos
Pessoal Técnico de Saúde , Tecnologia da Informação , Conhecimento , Humanos , Estudos Transversais , Tecnologia da Informação/estatística & dados numéricos , Radiografia , Sri Lanka , Inquéritos e Questionários , Pessoal Técnico de Saúde/psicologia
11.
Radiography (Lond) ; 29(1): 234-239, 2023 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-36608377

RESUMO

INTRODUCTION: Radiographer abnormality flagging systems have been in use in the UK for over 30 years, with the guidance of the Society and College of Radiographers indicated that the preliminary clinical evaluation (PCE), or comment, be the preferred system of choice. This study aimed to provide an updated assessment of current practice based upon a previous 2008 study. METHODS: A cross-sectional online survey was disseminated via Twitter and aimed at departmental and reporting leads. It requested information on the types of flagging and reporting systems operated, scope of the systems employed, required education of participants, and the role of audit. RESULTS: Responses were received from 31 Trusts within the UK. Red dot systems were employed in 90% (n = 28) of sites, with 26% (n = 8) undertaking PCE. Skeletal radiographs were most commonly reviewed (90%; n = 28) followed by chest (58%; n = 18) and abdomen (32%; n = 10). Only 13% (n = 4) sites indicated if the image was normal but 71% (n = 22) allowed the radiographer to indicate if they were unsure. There was marked variation in the educational requirements and use of audit. CONCLUSION: Compared to 2008 there appears to be quite minimal change in practices in the UK. There does appear to be some increase in the use of flagging systems generally and a higher proportion of PCE systems in comparison to red dot but the use of education and audit does not necessarily show much development in the past 15 years. IMPLICATIONS FOR PRACTICE: Significant conclusions cannot be drawn due to limited sample size, however, it may support further study and consideration in relation to implementation and potentially standardisation of abnormality detection systems may be justified.


Assuntos
Pessoal Técnico de Saúde , Controle de Qualidade , Humanos , Estudos Transversais , Radiografia , Reino Unido
12.
BMJ Case Rep ; 16(1)2023 Jan 02.
Artigo em Inglês | MEDLINE | ID: mdl-36593076

RESUMO

An adolescent girl came with swelling of both thighs with difficulty in walking for a few months. Based on clinical, and radiological evaluation and nerve conduction studies, she was diagnosed to be having bilateral hips Charcot's arthropathy due to hereditary sensory autonomic neuropathy type 4. Other common causes of Charcot arthropathy were ruled out. The patient was conservatively managed and parents were educated about preventive measures. Through this case report, we want to highlight the myriad number of manifestations and clinical presentations of a child presenting with hereditary sensory autonomic neuropathy type 4.


Assuntos
Artropatia Neurogênica , Neuropatias Hereditárias Sensoriais e Autônomas , Doenças do Sistema Nervoso Periférico , Feminino , Adolescente , Humanos , Criança , Doenças do Sistema Nervoso Periférico/complicações , Radiografia , Artropatia Neurogênica/diagnóstico por imagem , Artropatia Neurogênica/etiologia , Articulação do Quadril/diagnóstico por imagem
14.
Medicine (Baltimore) ; 102(3): e32730, 2023 Jan 20.
Artigo em Inglês | MEDLINE | ID: mdl-36701731

RESUMO

INTRODUCTION: Tuberculosis (TB) is an important infectious disease worldwide. Classical spinal TB has specific radiological findings involving adjacent vertebral bodies with destruction of the intervertebral disc and involvement of paravertebral soft tissues with cold abscess formation. However, a few cases not exhibiting the radiological characteristics of TB spondylitis have been reported. We report a rare case of lumbar spinal TB with atypical clinical and radiological presentations that was difficult to differentiate from a malignant spinal tumor. PATIENT CONCERNS: A 21-year-old man, who had immigrated to Japan from the Philippines 5 years ago, without a significant medical history, presented with back pain lasting 1 month and progression of gait disturbance 2 weeks prior to presentation. DIAGNOSIS: Laboratory tests showed normal blood cell counts and normal value of C-reactive protein levels. Preoperative imaging studies indicated a possible spinal tumor. However, histopathologic findings of the epidural soft tissues at the first surgery led to the diagnosis of spinal mycobacterial infection. The diagnosis of spinal TB was confirmed by a positive culture of Mycobacterium tuberculosis obtained at the second surgery. INTERVENTIONS: Given the progressive nature of neurologic deterioration, instead of needle biopsy, we proceeded with surgical intervention 8 days after admission; simultaneous neural decompression and open biopsy. Histological findings of the excised epidural soft tissues led to the diagnosis of spinal mycobacterial infection. We performed the second surgery involving additional resection of epidural soft tissues for further dural decompression and to obtain specimens for mycobacterial culture. Immediately after the second surgery, the patient commenced combination therapy with anti-tuberculous drugs. OUTCOMES: The patient demonstrated significant recovery of motor function in the lower extremities, and was able to run at 2 months after the second surgery. The epidural granulomas completely disappeared on magnetic resonance imaging 3 months postoperatively. CONCLUSION: Atypical clinical and radiological presentations of spinal TB present a challenge for appropriate diagnosis and early treatment. Even in developed countries where there are very few spinal TB patients, clinicians should be aware that spinal TB is an important differential diagnosis, especially in elderly patients or patients coming from countries with a middle-high prevalence of TB.


Assuntos
Mycobacterium tuberculosis , Neoplasias da Coluna Vertebral , Tuberculose da Coluna Vertebral , Humanos , Masculino , Adulto Jovem , Radiografia , Neoplasias da Coluna Vertebral/diagnóstico por imagem , Neoplasias da Coluna Vertebral/cirurgia , Coluna Vertebral/patologia , Tuberculose da Coluna Vertebral/diagnóstico por imagem , Tuberculose da Coluna Vertebral/cirurgia
15.
J Pediatr Orthop B ; 32(2): 170-177, 2023 Mar 01.
Artigo em Inglês | MEDLINE | ID: mdl-36700964

RESUMO

We report the rate of stable femoral condyle osteochondritis dissecans (OCD) lesion healing in adolescents using a locked hinged knee brace for a minimum of 6 weeks or until pain free, followed by physical therapy. Patients aged 5-18 at the time of a primary diagnosis of femoral condyle OCD were reviewed. Patients who presented with an unstable lesion necessitating surgery, lacked follow-up >6 months after diagnosis or before symptoms resolution, or were noncompliant were excluded. We assessed progression to surgery, association between surgery and lesion size, lesion grade, symptoms, laterality, and patient factors. Lesion width and depth and condyle size were measured on X-ray and MRI, and the scaled size of the lesion relative to the condyle was calculated. Sixty-four patients were included: 12.5% (n = 8; 50% female; mean age = 12.5 ± 1.0 years) progressed to surgery and 87.5% (n = 56; 20% female; mean age=11.5 ± 1.9 years) healed. No significant effects were identified between groups. The surgical group compared to the non-surgical group was braced for a similar amount of time (72.6 ± 51.4 vs. 54.9 ± 23.2 days; P = 0.09), presented initially with a similar lesion size (322.5 ± 298.7 vs. 211.2 ± 178.4 mm2; P = 0.14), and had a similar proportion of Grade 1 lesions (63% vs. 85%; P = 0.11). Female sex (P = 0.04) and longer time in the brace (P = 0.04) were associated with progression to surgery.


Assuntos
Braquetes , Fêmur , Articulação do Joelho , Osteocondrite Dissecante , Adolescente , Criança , Feminino , Humanos , Masculino , Fêmur/diagnóstico por imagem , Fêmur/patologia , Articulação do Joelho/diagnóstico por imagem , Articulação do Joelho/patologia , Osteocondrite Dissecante/diagnóstico por imagem , Osteocondrite Dissecante/terapia , Radiografia , Estudos Retrospectivos , Pré-Escolar , Resultado do Tratamento , Imageamento por Ressonância Magnética
16.
BMC Med Educ ; 23(1): 16, 2023 Jan 10.
Artigo em Inglês | MEDLINE | ID: mdl-36627640

RESUMO

BACKGROUND: Traumatic musculoskeletal injuries are a common presentation to emergency care, the first-line investigation often being plain radiography. The interpretation of this imaging frequently falls to less experienced clinicians despite well-established challenges in reporting. This study presents novel data of clinicians' confidence in interpreting trauma radiographs, their perception of AI in healthcare, and their support for the development of systems applied to skeletal radiography. METHODS: A novel questionnaire was distributed through a network of collaborators to clinicians across the Southeast of England. Over a three-month period, responses were compiled into a database before undergoing statistical review. RESULTS: The responses of 297 participants were included. The mean self-assessed knowledge of AI in healthcare was 3.68 out of ten, with significantly higher knowledge reported by the most senior doctors (Specialty Trainee/Specialty Registrar or above = 4.88). 13.8% of participants reported an awareness of AI in their clinical practice. Overall, participants indicated substantial favourability towards AI in healthcare (7.87) and in AI applied to skeletal radiography (7.75). There was a preference for a hypothetical system indicating positive findings rather than ruling as negative (7.26 vs 6.20). CONCLUSIONS: This study identifies clear support, amongst a cross section of student and qualified doctors, for both the general use of AI technology in healthcare and in its application to skeletal radiography for trauma. The development of systems to address this demand appear well founded and popular. The engagement of a small but reticent minority should be sought, along with improving the wider education of doctors on AI.


Assuntos
Inteligência Artificial , Músculo Esquelético , Médicos , Humanos , Computadores , Pessoal de Saúde , Radiografia , Sistemas de Apoio a Decisões Clínicas , Músculo Esquelético/diagnóstico por imagem , Músculo Esquelético/lesões
17.
RMD Open ; 9(1)2023 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-36609354

RESUMO

OBJECTIVES: To investigate whether in rheumatoid arthritis (RA) frequency of local joint inflammation is associated with radiographic joint damage progression in that joint. METHODS: Data from 473 patients with RA and available radiographs from the BeSt study were used. Patients were treated to target (Disease Activity Score of ≤2.4) for a median of 10 years. At each study visit every 3 months, joints were assessed for swelling and tenderness. Radiographs of hands and feet were made yearly. A generalised linear mixed model was used to assess the association between the percentage of study visits at which clinical inflammation was observed in a joint (cumulative inflammation) and radiographic joint damage in that same joint. Clinical inflammation was primarily defined as joint swelling (with or without joint tenderness). For secondary analyses, we also investigated joint tenderness without joint swelling. Damage was measured as the percentage of the maximum possible Sharp-Van der Heijde score in a particular joint. RESULTS: Cumulative local joint swelling was associated with local progression of radiographic damage in the same joint (ß=0.14, 95% CI 0.13 to 0.15). This association was also found in a subset of joints that were swollen at least once. Cumulative local joint tenderness without concurrent local joint swelling was less strongly associated with local radiographic joint damage progression (ß=0.04, 95% CI 0.03 to 0.05). CONCLUSIONS: In RA, long-term cumulative local joint inflammation is associated with joint damage progression in the same joint.


Assuntos
Artrite Reumatoide , Humanos , Artrite Reumatoide/complicações , Artrite Reumatoide/tratamento farmacológico , Artrite Reumatoide/diagnóstico por imagem , Inflamação/tratamento farmacológico , Radiografia , Edema , Quimioterapia Combinada
18.
Jt Dis Relat Surg ; 34(1): 69-74, 2023.
Artigo em Inglês | MEDLINE | ID: mdl-36700266

RESUMO

OBJECTIVES: This study aims to identify the most reliable level for the proximal locking screw in retrograde femoral nails and to investigate the preoperative detectability of the length of the proximal locking screw by radiological measurements. PATIENTS AND METHODS: Between April 2020 and June 2021, a total of 50 patients (42 males, 8 females; mean age: 38.1±14.3 years; range, 18 to 60 years) who were suspected of vascular injury after gunshot or stab wounds and underwent lower extremity computed tomography angiography (CTA) from the local institution's database were included in the study. The distances of the femoral neurovascular structures (FNVS) and sciatic nerve (SN) to the femur were measured in the sections determined in the anteroposterior and medial-lateral planes. The anteroposterior length of the femur was measured in selected sections to estimate the appropriate length of the proximal locking screw. RESULTS: The level at which FNVS and SN were closest to the femur in the medial-lateral plane was inferior to lesser trochanter (LT) 1 cm. The mean AP femur length at the level of the LT was 36.3±2.8 mm, at the level of inferior to LT 1 cm was 34.1±2.8 mm, at the level of superior to LT 1 cm was 38.6±3.7 mm. CONCLUSION: In retrograde femoral nailing, the safest level in terms of screw placement is 1 cm above the LT. Additionally, the optimal screw length is 40 mm at the level of the LT and 1 cm superior it, whereas it is 35 mm at the level of the LT and 1 cm inferior to it.


Assuntos
Fraturas do Fêmur , Fixação Intramedular de Fraturas , Masculino , Feminino , Humanos , Adulto Jovem , Adulto , Pessoa de Meia-Idade , Fixação Intramedular de Fraturas/métodos , Fraturas do Fêmur/diagnóstico por imagem , Fêmur/diagnóstico por imagem , Fêmur/cirurgia , Parafusos Ósseos , Radiografia
19.
Sci Rep ; 13(1): 1383, 2023 Jan 25.
Artigo em Inglês | MEDLINE | ID: mdl-36697450

RESUMO

Artificial intelligence (AI)-generated clinical advice is becoming more prevalent in healthcare. However, the impact of AI-generated advice on physicians' decision-making is underexplored. In this study, physicians received X-rays with correct diagnostic advice and were asked to make a diagnosis, rate the advice's quality, and judge their own confidence. We manipulated whether the advice came with or without a visual annotation on the X-rays, and whether it was labeled as coming from an AI or a human radiologist. Overall, receiving annotated advice from an AI resulted in the highest diagnostic accuracy. Physicians rated the quality of AI advice higher than human advice. We did not find a strong effect of either manipulation on participants' confidence. The magnitude of the effects varied between task experts and non-task experts, with the latter benefiting considerably from correct explainable AI advice. These findings raise important considerations for the deployment of diagnostic advice in healthcare.


Assuntos
Inteligência Artificial , Médicos , Humanos , Raios X , Radiografia , Radiologistas
20.
Crit Care ; 27(1): 40, 2023 01 25.
Artigo em Inglês | MEDLINE | ID: mdl-36698191

RESUMO

BACKGROUND: Chest radiographs are routinely performed in intensive care unit (ICU) to confirm the correct position of an endotracheal tube (ETT) relative to the carina. However, their interpretation is often challenging and requires substantial time and expertise. The aim of this study was to propose an externally validated deep learning model with uncertainty quantification and image segmentation for the automated assessment of ETT placement on ICU chest radiographs. METHODS: The CarinaNet model was constructed by applying transfer learning to the RetinaNet model using an internal dataset of ICU chest radiographs. The accuracy of the model in predicting the position of the ETT tip and carina was externally validated using a dataset of 200 images extracted from the MIMIC-CXR database. Uncertainty quantification was performed using the level of confidence in the ETT-carina distance prediction. Segmentation of the ETT was carried out using edge detection and pixel clustering. RESULTS: The interrater agreement was 0.18 cm for the ETT tip position, 0.58 cm for the carina position, and 0.60 cm for the ETT-carina distance. The mean absolute error of the model on the external test set was 0.51 cm for the ETT tip position prediction, 0.61 cm for the carina position prediction, and 0.89 cm for the ETT-carina distance prediction. The assessment of ETT placement was improved by complementing the human interpretation of chest radiographs with the CarinaNet model. CONCLUSIONS: The CarinaNet model is an efficient and generalizable deep learning algorithm for the automated assessment of ETT placement on ICU chest radiographs. Uncertainty quantification can bring the attention of intensivists to chest radiographs that require an experienced human interpretation. Image segmentation provides intensivists with chest radiographs that are quickly interpretable and allows them to immediately assess the validity of model predictions. The CarinaNet model is ready to be evaluated in clinical studies.


Assuntos
Aprendizado Profundo , Humanos , Traqueia , Intubação Intratraqueal/métodos , Radiografia , Unidades de Terapia Intensiva
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA