An accurate deep learning model for wheezing in children using real world data.

Auscultation is an important diagnostic method for lung diseases. However, it is a subjective modality and requires a high degree of expertise. To overcome this constraint, artificial intelligence models are being developed. However, these models require performance improvements and do not reflect the actual clinical situation. We aimed to develop an improved deep-learning model learning to detect wheezing in children, based on data from real clinical practice. In this prospective study, pediatric pulmonologists recorded and verified respiratory sounds in 76 pediatric patients who visited a university hospital in South Korea. In addition, structured data, such as sex, age, and auscultation location, were collected. Using our dataset, we implemented an optimal model by transforming it based on the convolutional neural network model. Finally, we proposed a model using a 34-layer residual network with the convolutional block attention module for audio data and multilayer perceptron layers for tabular data. The proposed model had an accuracy of 91.2%, area under the curve of 89.1%, precision of 94.4%, recall of 81%, and F1-score of 87.2%. The deep-learning model proposed had a high accuracy for detecting wheeze sounds. This high-performance model will be helpful for the accurate diagnosis of respiratory diseases in actual clinical practice.

Multi-axis shoulder abduction splint in acute burn rehabilitation: a randomized controlled pilot trial.

OBJECTIVE: The traditional thermoplastic aeroplane splint for axillary burn contracture is heavy, uncomfortable, and time consuming and difficult to put on and off. The present study tested the effectiveness of a newly designed multi-axis shoulder abduction splint with an easy-to-change angle. DESIGN: Randomized controlled parallel assessor blinded pilot. SETTING: Inpatient rehabilitation center in a general hospital. SUBJECTS: Twenty-four patients with recent (< 30 days) burns around the shoulder joint were randomized into two groups. After two dropouts, 11 patients used the new splint for four weeks and 13 patient were left unsplinted for four weeks. INTERVENTIONS: The newly designed multi-axis shoulder abduction splint keeps the shoulder abducted at the highest possible angle. MAIN OUTCOMES: The range of motion of the shoulder joint was measured at 0, 1, 2, 3, and 4 weeks. Active abduction, flexion, and external rotation were measured according to the zero position method by placing the axis of the goniometer ventral to the shoulder joint. RESULTS: Repeated-measure ANOVA revealed that the splint group developed significantly better abduction (P = 0.020) and flexion (P = 0.036) over 4 weeks than the non-splint group. ANCOVA using the initial (0 week) angle and Shoulder Burn Depth Index as covariates revealed that the splint group had significantly better abduction than the non-splint group (P = 0.013). CONCLUSION: The new multi-axis shoulder abduction splint resulted in a significant improvement in shoulder abduction angle compared to unsplinted patients.

Outcomes of ultrasound-guided extracorporeal shock wave therapy for painful stump neuroma.

OBJECTIVE: To investigate the effect of extracorporeal shock wave therapy (ESWT) on painful stump neuroma. METHODS: Thirty patients with stump neuroma at the distal end of an amputation site were assigned randomly to the ESWT group (n=15) and the transcutaneous electrical nerve stimulation (TENS)+desensitization+pharmacological treatment group (n=15). For 3 weeks, the ESWT group received a weekly session involving 1,500 pulses at 0.10 mJ/mm(2), while the control group was treated 10 times each, 40 minutes per day with TENS and desensitization treatment, and daily medication for 3 weeks. ESWT stimulation was given by focusing on the area at the neuroma site clearly identified by ultrasound. RESULTS: The changes in the McGill pain questionnaire were 38.8±9.0 prior to treatment and 11.8±3.1 following the treatment. The corresponding values for the control group were 37.2±7.7 and 28.5±10.3. The changes between groups were significantly different (p=0.035). The change in visual analog scale prior to and after treatment was 7.0±1.5 and 2.8±0.8 in the ESWT group, respectively, and 7.2±1.4 and 5.8±2.0 in the control group. These changes between the groups were also significantly different (p=0.010). The outcome in the pain rating scale also showed significant differences between groups (p<0.001). Changes in neuroma size and pain pressure threshold (lb/cm(2)) were not significantly different between groups (p>0.05). CONCLUSION: The study findings imply that ESWT for stump neuroma is superior to conventional therapy.

The effect of extracorporeal shock wave therapy on myofascial pain syndrome.

OBJECTIVE: To investigate the effect of extracorporeal shock wave therapy (ESWT) on myofascial pain syndrome (MPS). METHOD: Thirty patients with MPS in trapezius muscle were randomly divided into two groups, ESWT group (n=15), and trigger point injections (TPI)+transcutaneous electrical nerve stimulation (TENS) group (n=15). For a total of 3 weeks, ESWT was undertaken with 1,500 pulse each time at one week interval totaling 4,500 pulse, TPI for once a week totaling three times and TENS for five times a week totaling three weeks. RESULTS: The changes in pain threshold (lb/cm(2)) showed the values of 6.86±1.35 before first therapy, 11.43±0.27 after first therapy, and 12.57±0.72 after third therapy, while TPI+TENS group showed the values of 6.20±1.92 before first therapy, 8.80±0.48 after first therapy, and 9.60±2.19 after third therapy, and the changes between the groups were significantly different (p=0.045). The changes in visual analog scale were estimated to be 6.86±0.90 before first therapy, 2.86±0.90 after first therapy, and 1.86±0.69 after third therapy in case of ESWT group, whereas the figures were estimated to be 7.20±1.30 before first therapy, 4.60±0.55 after first therapy, and 2.80±0.84 after third therapy in case of TPI+TENS group, and the changes between the groups were significantly different (p=0.010). The changes in McGill pain questionnaire (p=0.816) and pain rating scale (p=0.644) between the groups were not significantly different. The changes in neck ROM were also not significantly different between the groups (p>0.05). CONCLUSION: The ESWT in patients with MPS in trapezius muscle are as effective as TPI and TENS for the purpose of pain relief and improving cervical range of motion.

The factors associated with contact burns from therapeutic modalities.

OBJECTIVE: To understand the injury pattern of contact burns from therapeutic physical modalities. METHOD: A retrospective study was done in 864 patients with contact burns who discharged from our hospital from January 2005 to December 2008. The following parameters were compared between patients with contact burns from therapeutic modalities and from other causes: general characteristics, burn extent, cause of burn injury, place of occurrence, burn injury site, treatment methods, prevalence of underlying disease, and length of hospital stay were compared between patients with contact burns. RESULTS: Of the 864 subjects, 94 patients were injured from therapeutic modalities. A hot pack (n=51) was the most common type of therapeutic modality causing contact burn followed by moxibustion (n=21), electric heating pad (n=16), and radiant heat (n=4). The lower leg (n=31) was the most common injury site followed by the foot & ankle (n=24), buttock & coccyx (n=9), knee (n=8), trunk (n=8), back (n=6), shoulder (n=4), and arm (n=4). Diabetes mellitus was associated with contact burns from therapeutic modalities; the odds ratio was 3.99. Injuries took place most commonly at home (n=56), followed by the hospital (n=33), and in other places (n=5). CONCLUSION: A hot pack was the most common cause of contact burns from therapeutic modalities, and the lower leg was the most common injury site. Injuries took place most commonly at home. The patients with contact burns from therapeutic modalities showed high correlation to presence of diabetes mellitus. These results would be helpful for the prevention of contact burns due to therapeutic modalities.

Effects of modified dynamic metacarpophalangeal joint flexion orthoses after hand burn.

OBJECTIVE: To assess the effectiveness of modified dynamic metacarpophalangeal joint flexion orthoses for treatment of post-burn hand contractures. METHOD: We enrolled 42 hand burn patients with limited range of motion at the metacarpophalangeal joints in this study. The patients were randomly assigned into either a control or an orthotic group. Both groups received the standard rehabilitation therapy focused on hand therapy; 21 subjects in the orthotic group wore a splint for 3 hours per day for 8 weeks. Hand function was measured by active range of motion, grip strength and other assessment tools. All parameters were estimated using the Mann-Whitney U test at the beginning and the end of the treatment after 8 weeks. RESULTS: The 21 subjects that had an orthotic intervention showed significant improvement in the range of motion at 2nd, 3rd, 4th and 5th metacarpophalangeal joints (p<0.05). However, the grip strength was not significantly increased after the 8 weeks of treatment compared to control group (p>0.05). There was a significant difference in the hand function scales between the 2 groups (p<0.05). CONCLUSION: The modified dynamic metacarpophalangeal joint flexion orthoses provide continuous flexion to metacarpophalangeal joint that is needed for the restoration of range of motion in post-burn hand contractures. For the clinical application of hand orthoses in patients with hand disorders, additional research into its affects are required.