Objective To establish a rapid, accurate and low-cost screening method for the detection of calreticulin (CALR) mutations in myeloproliferative neoplasms (MPN).Methods Seventy cases diagnosed with MPN were collected from 2012 to 2016. PCR combined with high resolution melting (HRM) analysis were used to screen the CALR mutations, and Sanger sequencing and T-A sequencing were applied to verify the HRM positive samples. CALR wild type DNA, type 1 and type 2 mutant DNA samples were selected and analyzed 4 times/day for 5 days to detected the CVs of Tm (melting temperature) respectively. JAK2 mutations were also analyzed in MPN patients to compare the association between JAK2 and CALR mutations.Results PCR-HRM analysis showed 7 cases (26.9%) and 5 cases (20.8%) patients with CALR mutations were screened out from 26 essential thrombocythaemia (ET) cases and 24 primary myelofibrosis (PMF) cases, but no CALR mutations were found in cases with polycythaemia vera (PV). All mutations were confirmed by direct sequencing or cloning sequencing. The CVs for HRM analysis of CALR wild type DNA, type 1 and type 2 mutant DNA samples were 1.91%,1.59% and 1.43%, respectively.There were 47 cases with JAK2 V617F and 1 case with exon12 mutation. No coexistence of JAK2 mutation and CALR mutations were found in a single sample.Conclusion PCR-HRM can be used for rapid screening of CALR mutation. Subsequent sequencing can be applied for rapid diagnosis of MPN patients in clinical practice.

BACKGROUND:More and more studies employ CD44 as a specific marker of colorectal cancer stem cells. Oct4 is a transcription factor of embryonic stem cells, and it has been discovered recently that there is a higher expression in primary colorectal carcinoma.
OBJECTIVE:To investigate the quantity, location and distribution of CD44+/Oct4+cells in primary colorectal carcinoma.
METHODS:A total y of 108 cases of human colorectal carcinoma and 18 cases of normal mucosa, 18 cases of adenoma were col ected and made into three tissue microarrays, each containing of 48 dots. The locations of CD44+/Oct4+cells were detected by double-label immunohistochemical staining and hematoxylin-eosin staining. The morphologic features of them were investigated on hematoxylin-eosin staining at the same position. 
RESULTS AND CONCLUSION:The results of double-label immunohistochemical staining demonstrated that there were no CD44+/Oct4+cells in normal intestine mucosa and a very smal amount of CD44+/Oct4+cells in adenoma, and double-positive cells could also be seen in colorectal carcinoma. The number of CD44+/Oct4+cells was rare and the cells were scattered or distributed focal y along the basement of gland basal side. The cells with scarce cytoplasm were square, and its nucleus was oval or high cylindrical, deeply stained and homogeneous. The quantity of CD44+/Oct4+cells was negatively correlated with the differentiation of colorectal cancer (r=-0.579, P<0.01), and was associated with the depth of tumor invasion (r=0.236, P<0.05). These findings indicate that CD44+/Oct4+cells may be colorectal cancer stem cells.

Diabetes diet management plays an important role in the treatment of diabetes. "Controlling diet" is the most basic and important part of diabetes treatment. Patients with mild diabetes can control blood glucose through diet therapy. Effective diet management assessment can quickly discover the deficiencies of diet self-management in diabetic patients. Artificial intelligence is widely used in the medical field. This article will briefly introduce the role and application progress of artificial intelligence technology in diabetes diet management, including diet recommendation and automatic monitoring.