Artificial intelligence (AI) is evolving at a tremendous pace and is increasingly permeating the medical field. In the field of medical image diagnosis, the performance of so-called CAD (computer-aided diagnosis) systems that support doctors’ image diagnosis has significantly improved as AI-CAD due to the recent emergence of deep learning, the core technology of AI. AI products are being introduced into medical settings, with some even offering insurance assistance. Such AI is classified as “discriminative AI (recognition AI),” but with recent “generative AI,” AI-CAD will likely evolve into generative AI-CAD that is even closer to that of human doctors. This review provides an overview of the current state of AI utilization in medical care, explains the basics of deep learning and generative AI (large-scale language models, Transformer, ChatGPT, image generation AI, basic models, etc.), which are technologies that are attracting attention as a driving force, and then discusses the current status and challenges of introducing image diagnosis AI into clinical settings.

  Objective: While the Hiragana reading test is essential for the diagnosis and treatment of developmental dyslexia in Japan, the current norm may overdiagnose reading difficulties, especially among lower graders. The present study aimed to verify the current norm and the extent of overdiagnosis based on a large sample of first graders (G1) and second graders (G2). In addition, more appropriate norm was explored using longitudinal data. Methods: A total of 511 and 422 students enrolled in G1 and G2 at public elementary schools, respectively, completed the Hiragana reading test. First, we compared the reading time distributions between the present sample and the current norm. Second, the diagnosis rate of reading difficulty was examined for each grade. Finally, a new norm was established for the Hiragana reading test based on a three-year longitudinal data, which were contrasted with the current norm with respect to diagnosis rate and longitudinal trajectory. Results: Reading times in the present G1 and G2 samples were considerably longer than the current norm. Diagnosis rate of reading difficulty reached as high as 47.0% in G1 and 30.3% in G2. The longitudinally established new norm was more moderate and led to diagnosis rate of around 7% for both grades. In a comparison of three-year trajectory, nearly half of the students were classified as having reading difficulties, at least once based on the current norm, whereas the use of new norm dropped the classification rate to around 15% and improved overdiagnosis. Conclusions: The current norm of the Hiragana reading test was overly strict for both G1 and G2 graders, which indicates that there is a high risk of overdiagnosing reading difficulties. The new norm was suggested to be more useful for both medical diagnosis and educational intervention. Therefore, it is necessary to modify grade norms of the Hiragana reading test.

  A 7-years-old girl was admitted with the principal complaint of abnormal gait. She presented with distal muscle weakness, bilateral Achilles tendon reflex depression, and peripheral sensory disturbance. The results of the spinal cord magnetic resonance (MR) neurography were normal ; however, she had median nerve conduction disorders. As a genetic analysis revealed a mitofusin 2 mutation (c.718T＞G, p.F240V), she was diagnosed with Charcot-Marie-Tooth (CMT) type 2A. Despite continued physical therapy, her symptoms progressed slowly. At the age of 10 years and 4 months, she was unable to walk without support, and spinal MR neurography demonstrated bilateral nerve root enlargement in L2-5 regions. In children with CMT type 2A, the morphology of the spinal cord nerve root is not abnormal in the early stages, but nerve root enlargement may be observed as symptoms progress. Spinal MR neurography may be useful in evaluating disease activity.

  Pallister-Killian syndrome (PKS), a sporadic congenital anomaly syndrome associated with intellectual disability, epilepsy, movement disorders, and multiple organ malformations such as diaphragmatic hernia, is caused by mosaic tetrasomy of the short arm of chromosome 12 (12p). Acute gastric volvulus is an uncommon but fatal gastrointestinal disorder often caused by abnormalities in the ligaments that support the stomach and can occur as a secondary condition to diaphragmatic hernia. We experienced a case of acute gastric volvulus in an adult patient with PKS and adult-onset chronic Bochdalek hernia. The patient was diagnosed with PKS at 11 years of age and has been institutionalized since she was 13 years old. The patient was diagnosed with a chronic Bochdalek diaphragmatic hernia at 33 years old. When the patient was 36 years old, she presented with a distended upper abdomen and was in hypovolemic shock. During the evaluation, the patient did not vomit, and we could not pass a nasogastric tube. The patient was diagnosed with gastric volvulus using radiography and abdominal plain computed tomography and was transferred to an emergency hospital 5 hours after detecting the change in her condition. Blood analysis performed at the emergency hospital revealed severe acidosis and hyperkalemia due to abdominal compartment syndrome caused by the gastric volvulus. Emergency exploratory laparotomy revealed that the stomach was necrotic ; upon decompressive incision of the dilated stomach, the patient developed ventricular fibrillation. An advanced cardiac life support protocol was attempted without resuscitation. Gastric volvulus should be taken into consideration in patients with congenital conditions, such as PKS, presenting with acute upper abdominal distention. Gastric volvulus is a surgical emergency that should be treated promptly. If a diaphragmatic hernia is detected in patients with congenital conditions associated with intellectual disability, elective surgery should be considered even if hernia-associated symptoms are not obvious.

  We report a case of anti-N-methyl-D-aspartate (NMDA) receptor encephalitis in a 2-year-old boy. He presented with frequent partial seizures in the left side of his body with no preceding infection or fever, followed by mild left-side hemiplegia and gait disturbance. During the course of the seizure, stereotypic, lip licking movement of the tongue was observed. Also, he developed irritability and night delirium at the eleventh day from the onset. Initially, he was treated as epilepsy and received an anti-epileptic drug because his level of consciousness in the daytime was normal and his magnetic resonance imaging scan revealed no abnormalities. Eighteen days later, he presented with status epilepticus, accompanied by persisting impaired consciousness. He received methylprednisolone pulse therapy, high-dose immunoglobulin therapy, and thyrotropin-releasing hormone therapy. Although hemiplegia resolved and no motor sequela was observed, higher brain dysfunction remained. With identification of anti-NMDA receptor antibody, he was diagnosed with anti-NMDA receptor encephalitis 2 months after the initiation of symptoms. Adult patients of anti-NMDA receptor encephalitis often show psychiatric symptoms initially. However, in case of pediatric patients, seizure and involuntary movement are reported as initial symptoms. Although in retrospect, the clinical course of our patient was typical of pediatric anti-NMDA receptor encephalitis, it was not easy to reach the final diagnosis. We suggest that proper knowledge of disease characteristics of pediatric anti-NMDA receptor encephalitis is essential for early diagnosis and treatment. Also, clinical diagnostic criteria that do not rely on time-consuming antibody testing are suggested. Applying such criteria to pediatric cases might be beneficial for early diagnosis.

  A 9-year-old boy who was taking multiple medicines for ADHD and autism spectrum disorder presented to our hospital. He suffered from uncontrollable tongue protrusion and involuntary movement of the lower jaw after he discontinued taking extended-release methylphenidate on his own. His symptoms disappeared spontaneously, but he had experienced two similar episodes previously. We diagnosed him with dystonia due to discontinuation of extended-release methylphenidate during concomitant use of risperidone. Precaution should be taken while changing the drugs’ dose when ADHD stimulants and antipsychotics are used concomitantly.

  We present a case of molybdenum cofactor deficiency type A (MoCD-A) which was asymptomatic in the neonatal period but developed in infancy, displaying motor developmental delay and a clinical course resembling Leigh’s encephalopathy. There was no evidence of hypouricemia or abnormalities in generally recognized serum amino acid levels (taurine, cystine, etc.). A previously reported homozygous variant of the MOCS1 gene (NM_001358530.2 : c. 949C＞T [p. Arg317Cys]) was identified by whole-genome sequencing, leading to the MoCD-A diagnosis. This disorder should be considered in patients with Leigh encephalopathy-like clinical manifestations.