Early detection of multidrug-resistance mutations is critical to the effectiveness of antimicrobial treatments in infection diseases. Culture-based antimicrobial susceptibility test is highly recommended before antibiotic regimes. However, this method is not suitable for fastidious bacteria such as Mycobacterium tuberculosis. For slow growth pathogens, molecular tests with the amplification of target sites are widely used in clinical settings. These tests possess limitations on the well-known positions of resistance mutations and resistance genes, meaning that if resistance mutations occur outside these locations, they will be missed. Next generation sequencing technologies have made them feasible for detecting current resistance mutations as well as monitoring the emergence of new mutations. Short read sequencing is currently the most commonly form and has wide range of diagnostic applications. However, it requires complexity settings that make it unsuitable for remote areas.Moreover, its constraints in read length prevent the sequencing of long stretches of DNA. Long read sequencing such as Oxford Nanopore Technologies’ platform capable of producing reads of up to 1 million base pairs by using portable devices. A significant challenge of the technologies lies in the low bacilli abundance and high human tissue in clinical samples , resulting in insufficient sequencing depth. To incorporate long read sequencing technologies to clinical testing, we develop a rapid diagnostic test to detect multi-drug resistance direct from clinical samples with low bacilli abundance such as human sputum. We generate magnetic particles to specifically capture the target organism. Whole-genome amplification is applied by using specific primers to further enrich the target genomes. We then apply the test to sputum samples to detect multi-drug resistance M. tuberculosis (MTB). We are able to capture most of MTB cells in the spiking samples using the nanoparticles. The specific primers increase MTB DNA concentration more than 100 folds in mixed samples of high human and low MTB DNA. The success of the rapid test will contribute to improve antibiotic treatments and monitor the emergence of new resistance mutations in clinical settings especially in low-income countries.